The Role of Emotions in the Capabilities Approach: A Critical Analysis

Abstract

The capabilities approach is the theory according to which, in order to assess people's quality of life and reflect on the basic political entitlements, we should consider what people are capable of doing and being. Focusing mostly on Nussbaum's account, a number of scholars analysed the metaethical structure underlying the approach, showing her Aristotelian and Kantian sources. This article explores another aspect of Nussbaum's theory which has so far been somewhat overlooked: the role of emotions in the justification and motivational support of the approach. After showing the importance Nussbaum places on compassion (and other benevolent emotions) to warrant her moral theory, I argue that – assuming her own cognitive account of compassion – her argument begs the question. I will then contend that Nussbaum's case does not have the inner tools to overcome such a circularity, trying to suggest, in conclusion, a possible way out which is consistent with it.

1. Introduction

The capabilities approach is one of the most promising and discussed theories of justice of recent years. Its key claim is that, to inquire into pivotal moral-political questions concerning welfare and respect for the dignity of citizens, it is not enough to focus on what people possess (i.e., the resources at their disposal), but it is necessary to assess what they are concretely capable of doing and being. Precisely because of its ambitions, also recognised by its opponents, the capabilities approach has drawn the attention of many scholars, who have analysed and challenged its theoretical foundations.

Metaethical investigations have focused above all on Martha Nussbaum's account, which makes more extensive use of substantive philosophical concepts. Scholars who have dealt with these issues have mostly emphasised two different sources. On the one hand, many have conceived Nussbaum's theory as a form of Aristotelianism. A number of authors (Wallach, 1992; Bobonich, 1993; Alexander, 2005) have underlined the role of concepts such as human nature, characteristic functioning, and the conception of practical deliberation underlying her proposal, drawing comparisons with the internal realism of Hilary Putnam and other neo-Aristotelian approaches (sometimes even rather different from the liberal one of Nussbaum) such as that of Alasdair MacIntyre (Zuckert, 2014). Michele Moody-Adams (1998) has defined Nussbaum's approach as ‘Aristotelian essentialism’, while Den Uyl and Rasmussen (2009) criticised her Aristotelian concept of human flourishing, raising doubts about its capacity to justify her political conclusions. On the other hand, others have also underlined the influences of Kantian prescriptivism. For example, Paul Formosa and Catriona Mackenzie (2014, pp. 882–87) compared Nussbaum's notion of dignity with that of Kant, underlining the similarities between them. Likewise, Rutger Claassen and Marcus Düwell (2013), showing how Nussbaum borrows from Kant concepts such as human dignity or the idea of treating individuals as ends, compared her model of justification of the central capabilities with Alan Gewirth's neo-Kantian account. More precisely, they (2013, pp. 493–95, 506–9) argue that Nussbaum's universalistic and categorical claims would have a more sound foundation through a ‘transcendental justification’ of capabilities, i.e., showing that there are capabilities that are necessary for people not only to achieve a good life but to even consider themselves practical agents.

Of the three leading metaethical paradigms usually acknowledged (Appiah, 2003), the only one to which the capabilities approach has never been compared is the so-called ‘emotivism’. In some respects, this is not surprising. According to emotivism, moral sentences have no truth value (they are neither true nor false) but are mere expressions of feelings, personal preferences, and desires; and, although it is not necessary, such a view is often associated with relativism. Conversely, the capability approach in general (and Nussbaum's version in particular) aims at providing an objective account of well-being or flourishing. Nussbaum's harsh rebuttal of the most radical forms of relativism, as well as her critique of preferences as an adequate basis for assessing people's well-being, represents a major difference from that metaethical view (Nussbaum, 2000, pp. 34–110; 1999, pp. 29–54, 118–29). Hence, I will not try to force such a bold comparison. However, there is an element of Nussbaum's account which, albeit with due differences, brings it closer to emotivism and which, surprisingly, has not yet been adequately taken into account: the importance of emotions for justifying and supporting her capability theory. This lack is even more puzzling considering that the original theory of emotions she developed is widely discussed (Nussbaum, 2001a).¹

In what follows, I will try to fill this gap, showing the emotional assumptions underlying Nussbaum's proposal and suggesting a possible flaw. In sections 2 and 3, I will summarise Nussbaum's version of the approach, emphasising the role of benevolent emotions – above all compassion – in supporting her political proposal. In sections 4 and 5, I will focus on Nussbaum's theory of emotion (in particular, on her cognitive account of compassion), showing that, assuming the truth of both of her theses (the moral-political one on the importance of benevolent emotions for her capabilities approach and the psychological one on their cognitive bases), she ends up begging the question. In the conclusions (section 6), I will consider a possible response to my critique – indirectly deductible from Nussbaum's writings – arguing that it too fails to solve the circularity. Finally, I will try to sketch a different solution, but still consistent with her ethical theory.

2. Martha Nussbaum's Capabilities Approach

The capabilities approach is the view according to which, to correctly assess people's quality of life (and make comparisons for economic or political purposes), we should consider what people are capable of doing and being, rather than relying on the goods they have or the satisfaction of personal preferences.² Indeed, a purely economic evaluation of the quality of life (which considers only the resources people possess) has two major problems. First, it fails to grasp all those values, central to a human existence, beyond the economic realm, such as average life expectancy, the opportunity to receive adequate education, or political freedoms and civil rights. Second, models focused just on primary goods overlook the fact that not all of us are equally capable of ‘converting’ the same number of economic resources into actual functioning. For instance, knowing that a person in a wheelchair has a certain income does not ensure that they are able to move around as freely as an able-bodied person can (Nussbaum, 2000, 2006; Sen, 1982, 1992).

On the other hand, utilitarian approaches that focus on satisfying individual preferences fail to recognise those cases in which being used to a condition of need inhibits people's own desires. Nussbaum (2000, p. 111) speaks of ‘adaptive preferences’ to indicate this phenomenon: people who live in prolonged plight and deprivation tend to reshape their preferences downwards, resigning themselves to their condition; in this way, it is possible for someone to stop desiring a basic good for the simple fact that they know it is impossible for them to obtain it. For example, some women in India consider it normal – hence, in some sense, satisfactory – being denied access to education, clean water, etc., equal to men. Analysing the quality of life starting from subjective preferences (without assessing whether or not they are ‘informed’) makes it more difficult to be aware of these social distortions of desire.³

Conversely, the capabilities approach reflects on the well-being of individuals starting from their ‘capabilities’ and ‘functioning’. The definitions of these terms among capability scholars are diverse and debated.⁴ Some of them refer to capabilities as ‘real freedom’ (Byskov, 2020) or ‘substantive opportunity’ (Byskov, Kramm and Östlund, 2020) to be or do something. However, as Brukamp (2001) has underlined, in Nussbaum's sense they can be better understood with reference to the Aristotelian distinction between act and potency. Functionings represent individuals’ concrete ways of being or acting, what they are and what they do – in other words, the actual ways of performing one's functions: being well nourished or not, taking part in public life or not, etc. Capabilities, on the other hand, define one's own potential, what one can be and what one can do, or rather their possibility of functioning in certain desired ways, pursuing the lifestyles they find fulfilling (Nussbaum, 2000, pp. 70–74; Nussbaum, 2011, pp. 20–25).⁵

Nussbaum develops these concepts based on other Aristotelian intuitions. She notes there are functions that contribute to defining human existence in a dual sense: ‘first, […] certain functions are particularly central in human life, in the sense that their presence or absence is typically understood to be a mark of the presence or absence of human life; and second […] there is something that it is to do these functions in a truly human way, not a merely animal way’ (Nussbaum, 2000, pp. 71–72) – that is, a way permeated with practical reason and sociability. There are at least two Aristotelian insights underlying this account. First, what Alasdair MacIntyre (1981, p. 58) calls a ‘functional concept’ of human being: just as we define a watch by its characteristic function of measuring time (such that, if it did not measure time, it would not be a real watch), in the same way, for Aristotle, there are characteristic functions that define a fully human existence (or, as Nussbaum claims, a life worthy of a human being).⁶ It is by freely exercising these functions that we can live a good life. According to Aristotle, the essential functions of human beings are a) rationality and b) sociability. For Nussbaum, the list of human functions is more extensive (as we will see, it includes ten items). But – this is her second Aristotelian debt – the two that ‘stand out as of special importance, since they both organize and suffuse all the others’ (Nussbaum, 2000, p. 82) overlap the Aristotelian ones (Brukamp, 2001, p. 99): she labels them as ‘practical reason’ and ‘affiliation’.

If we recognise that a life worthy of a human being is characterised by the free exercise of some pivotal functions, respect for dignity requires the state to guarantee each citizen the capacity to function (if one so wishes) in the fields of experience that are central to human fulfilment. For this reason, Nussbaum proposed a list (deliberately open and generic, to allow convergence even by people who support very different global conceptions of a good life) of ten central capabilities, which define the social minimum every state should ensure its citizens so as to offer them the opportunity to enjoy a decent life. In its most updated version, the list includes: 1) Life; 2) Bodily Health; 3) Bodily Integrity; 4) Sense, Imagination, and Thought; 5) Emotions; 6) Practical Reason; 7) Affiliation; 8) Other Species; 9) Play; 10) Control over One's Environment, both political and material (Nussbaum, 2000, pp. 78–80; Nussbaum, 2006, pp. 76–78; Nussbaum 2011, pp. 33–34).⁷

3. The Role of Compassion for the Capabilities Approach: Beyond Contractarianism

We have seen Nussbaum's debt to the Aristotelian concepts of proper functions and a fully human life. On the other hand, it is possible to read the central capabilities also as constraints demanding unconditional respect, like the Kantian imperative.⁸ Furthermore, Nussbaum herself places her approach within the Rawlsian tradition of political liberalism, insisting on the need to reach an overlapping consensus on the list, beyond people's own global conceptions of the good. (I do not mean to discuss here the widely debated question (Arneson, 2000; Richardson, 2000; Biondo, 2008; Katzer, 2010; Carter, 2014) of whether Nussbaum's theory is truly liberal or a form of perfectionism.) However, it seems to me that those who have dealt with this issue have so far overlooked the central role that Nussbaum assigns to emotions, and in particular to compassion, in her account. This can clearly be seen in Frontiers of Justice (Nussbaum, 2006), where she compares her theory with the ‘rival’ tradition of contractarianism, arguing that her approach is better suited to respond to three problems of contemporary ethics: a) the just treatment of persons with disabilities, b) transnational justice, and c) the rights of non-human animals. In all these cases, Nussbaum argues, contractarianism fails to support a number of claims that her theory, by contrast, considers as pivotal.

The main source of this difference between Nussbaum's capabilities approach and contractarianism lies precisely in their different anthropological and ‘emotional’ assumptions. According to her, the main flaw that dooms contractarianism to failure in the three ‘frontiers of justice’ analysed is the fact of thinking of justice as a contract between individuals – free, independent and almost equal in terms of rational capacities – who agree on fundamental principles for mutual advantage. For Nussbaum (2006, pp. 9–95), this assumption entails two problematic forms of exclusion. First of all, it mistakenly identifies the recipients of (i.e., those who are protected by) the principles of justice with those who plan them rationally. In doing so, all subjects devoid of faculties considered essential to subscribe to the social contract – such as equal rationality, autonomy, independence – are excluded ex ante from it. Insofar as they cannot be included in the group of contractors who develop the principles of justice, both animals and people with severe cognitive disabilities cannot, properly speaking, even be subjects of justice. Secondly – and this is the most relevant fact for our purposes – even if contractarians were able to address this problem, there would still be a problem with another of their key assumptions: the idea that the social agreement is grounded on the fact that, in this way, the contractors will gain a mutual advantage. If the protection of one's own interest is the only motivation for entering into a collective agreement for social cooperation, Nussbaum argues, it is unclear why we should include much weaker categories, with whom reaching a fair agreement is not at all advantageous. This applies both to the aforementioned cases of people with mental disabilities and animals, and to transnational justice, i.e., our duties towards citizens of other nations (especially the underprivileged ones). In all these cases, a contract would be unfavourably demanding for the stronger party (human beings without disabilities, citizens of prosperous nations), who would be supposed to make significant sacrifices not rewarded by equal benefits.

Contrary to contractarianism, the capabilities approach presupposes a richer conception of the motivation underlying the search for justice, which is not only based on self-interest but also includes ‘a moralized compassion for those who have less than they need to lead decent and dignified lives’ (Nussbaum, 2006, pp. 156–57; cf. pp. 85–92, 156–59, 222–23, 323–24).⁹ From this point of view, human beings are led to cooperate following fair rules not only for their own advantage but due to a desire for justice and a set of benevolent emotions, that is, an interest in everyone having the opportunity to flourish. Although Nussbaum does not make the passages underlying her argument so explicit, we could summarise them as follows. The basic assumption is the idea that a flourishing life consists in the free exercise of some capabilities, characteristic of human existence. A supporter of this approach will acknowledge that a life devoid of these capabilities is not worthy of human dignity, preventing the person from realising a good life (eudaimonia). But, if they accept this, they could not be indifferent to those who, unfairly, lack them. Such a situation will elicit the supporters’ compassion, which will provide them with the psychological motivation to help also the victims to have the (political and social) opportunity to live a decent existence. I will call this line of argument the case for moral compassion.

If my account of the reasons that move the ideal supporter of the capabilities approach is correct, however, Nussbaum has a problem: the moral psychology underlying her political theory is inconsistent with some of her claims concerning the structure of emotions (more precisely, concerning the structure of compassion). To understand this problem, we need to briefly consider her cognitive theory of emotion, and analyse her definition of compassion. I will sketch both in the next section, and then show, in section 5, how they conflict with Nussbaum's claim about the role of compassion in grounding the capabilities approach.

4. Nussbaum's Cognitive Account of Compassion

Although Nussbaum claims a central role for compassion (and other benevolent emotions) in grounding her theory of justice, her view is not a romantic and uncritical form of sentimentalism, nor a care ethics that maintains the ‘wisdom of the heart’ against rational principles of fairness. Rather, in a number of works, she argues that it is necessary to combine two equally important components in mutual support: on the one hand, an appropriate ability to feel empathy and compassion for others; on the other, fair principles, i.e., a rationally justified theory of justice (Nussbaum, 1995; 2013).

Besides being a rather reasonable position from a practical point of view, it is perfectly consistent with the definition of emotion Nussbaum outlined extensively in her masterpiece Upheavals of Thought (2001a, in particular pp. 33–56). In it, she defends a ‘cognitive-evaluative theory’, according to which emotions are not blind and irrational energies, but are closely intertwined with the subject's beliefs. Unlike appetites such as hunger and thirst (which are impulses aimed at satisfying one's physical needs), emotions have a cognitive content and embody evaluative judgments about the world: in experiencing them, we judge a fact or an event as positive or negative, helpful or harmful, with reference to our most pivotal goals and ends. With a felicitous expression, Nussbaum states that emotions are eudaimonistic, to mean that they express a judgment on the importance of certain external goods (people, situations or objects) for one's own happiness, or flourishing. (Psychologists, instead, use the term appraisal to denote the same feature, namely that emotions represent a form of evaluation – however quick and unreflective – of the stimulus that produced them.) So, for example, fear is not just a thrill or a heartbeat (which may be present or not); it consists in the thought that something important to our well-being is in danger. Similarly, grief is not just a gut feeling, but the awareness that someone important to us is lost forever.

Hence, emotions are cognitive phenomena that imply that the subject has certain beliefs, not necessarily expressed or expressible in propositional form, i.e., elaborated as complete sentences, as if we were to pronounce or write them down. Rather, very often they are unreflective thoughts, to which we do not pay much conscious attention – but they are beliefs nonetheless. This is worth noticing. For, if emotions have cognitive content, they can be proper objects of rational assessment (Deigh, 1994; Nussbaum, 2001a). This means that, for a cognitivist, they can be more or less rational, according to the truth of the beliefs underlying them. I do not mean to address here the already widely discussed question of whether Nussbaum's theory is a reliable account of emotion or not.¹⁰ What I want to argue is that, taking for granted her own account of emotions (and of compassion in particular), it raises problems for the justification of the capabilities approach I sketched in section 3.

According to Nussbaum, compassion – like all other emotions – is based on some thoughts, or judgments, of the subject who experiences it. More precisely, borrowing some of Aristotle's intuitions, she argues that compassion rests on three essential beliefs, which she labels respectively as ‘judgment of size’, ‘judgment of nondesert’ and ‘eudaimonistic judgment’. The first ‘cognitive requirement’ concerns the effective gravity of the situation: ‘We do not go around pitying someone who has lost a trivial item, such as a toothbrush or a paper clip, or even an important item that is readily replaceable’ (Nussbaum, 2001a, p. 307; cf. 304–27). On the contrary, we pity those who have suffered serious losses, which affect their capability to flourish. (This does not mean, as some have objected (Carr, 1999), that for Nussbaum we pity others only for those losses that would make us feel bad too in the first person: for instance, we may pity a trombone player for a little injury on his lip, because this would prevent him from playing in an important concert, although – personally – such a wound would have caused us only minimal discomfort. However, the reason we feel compassion is that we judge the situation as serious, not trivial, based on our own conception of eudaimonia, according to which work fulfilment is an essential element for human flourishing – even if, maybe, for us this fulfilment does not imply playing an instrument but managing a successful business or editing a special issue of a prestigious journal.¹¹)

The second cognitive element of compassion for Nussbaum is the judgment of nondesert, i.e., the idea that the sufferers do not deserve what happens to them. For example, according to her, we are unlikely to pity a wealthy American who cries at the news of a sudden tax hike because he has already spent that money on a lavish ski vacation in Aspen. In such a case, Nussbaum (2004, p. 483) argues, compassion would be inhibited by the idea that the wealthy taxpayer would find himself in that situation only because of his pretentious lifestyle. Finally, according to Nussbaum, compassion also entails a third belief, which she calls the ‘eudaimonistic judgment’. Like other emotions, to even feel compassion it is necessary to consider the object of the emotion (in this case someone's undeserved plight) ‘as a significant part of his or her own scheme of goals and ends’; that is, one ‘must take that person's ill as affecting her own flourishing’ (Nussbaum, 2001a, p. 319). Since we usually struggle to develop such a strong attachment to strangers, the eudaimonistic judgment may be eased by another belief, which Aristotle considered necessary for compassion: what Nussbaum calls the ‘judgment of similar possibilities’, i.e., the idea that we could ourselves suffer from a harm akin to the one which now affects the victim. While Aristotle held this belief to be a necessary condition for compassion, Nussbaum (2001a, pp. 315–21) does not consider it essential. For, while acknowledging that this thought often helps us to be involved in other people's sufferings, she argues it is not contradictory to imagine an invulnerable divine being – such as the Judeo-Christian God – who nevertheless pities mortals’ plights.

4.1 Flaws of Compassion and How to Fix Them

While arguing for the moral value of compassion and denying the idea that human beings aim always to maximise their own self-interest, Nussbaum (2001a, pp. 414–25) is aware that there are many ways in which compassion can fail. In her cognitivist perspective, these flaws stem from errors of evaluation in one of the three essential judgments underlying the emotion (that of size, of nondesert, and the eudaimonistic one). For instance, it is possible for someone to fail to pity a real plight, such as the loss of civil rights or the impossibility of receiving education, because they do not consider these goods pivotal to a flourishing life. Other times, our compassion is inhibited due to a biased judgment of nondesert, as is the case with many Americans who think of poverty as the victim's fault. But, for Nussbaum (2001a, p. 420, original emphasis), ‘[t]he judgment that goes wrong most often, and most dramatically, is the judgment of the proper bounds of concern, or what I have called the eudaimonistic judgment’. Frequently, we fail to consider others’ well-being (at least in part) as important to our flourishing. We tend to feel more sympathy for those who look like us, or with whom we share social status, or those who stand before our eyes. Conversely, we struggle to pity those who are more socially or spatially distant. Furthermore, some negative emotions, such as fear or disgust, can further narrow our ability to take an interest in the conditions of strangers, making them appear to us as not fully human or focusing one's attention exclusively on one's own well-being and survival.¹²

Insofar as our natural compassion has these flaws, it cannot ground morality alone. This is why Nussbaum suggests refining it through philosophical and political work. In fact, following the cognitivist perspective, in all these cases, the inappropriate emotional response depends on the subject's false beliefs. But they can be revised through rational reflection and habit. Thus, according to Nussbaum (2001a, p. 392; cf. pp. 414–25), the ‘solution to its partiality problems is to work on compassion's developmental history, trying to get the three judgments right through appropriate education and institutional design’. As she puts it in Political Emotions (2013), where she reflects on the role emotions should have in our liberal democracies,

we must not regard compassion as an uncriticized foundation for public choice. Emotional foundationalism is as pernicious as neglect of the emotions. […] We must arrange for a continual, and watchful, dialogue between vivid imagining and impartial principle, seeking the best and most coherent fit, always asking what we're entitled to give to those whose situation we vividly imagine and how far we need, by contrast, to follow impartial principle. (Nussbaum, 2013, p. 157)

Hence, Nussbaum's proposal is to let both philosophical reflection and political institutions lead the citizens’ beliefs, so that they feel (morally) appropriate forms of compassion – i.e., a compassion based on three true judgments. Here are some examples of how philosophy and just institutions can reshape people's minds: by protecting some basic rights, constitutions can inform citizens’ judgment of size, indirectly suggesting that some features are key for human dignity and their loss is worthy of compassion. For instance, the fact that the Constitution protects free speech inclines people to think that its loss would be something to pity. Institutions can also guide and shape the judgment of nondesert: a welfare state can prevent people from thinking of poverty as the result of the subject's idleness or lack of boldness. Similarly, it is plausible that changes in rape law have shaped public opinion over time, allowing them to overcome the idea that the women themselves provoke the assault because of their clothing or incautious behaviour (Nussbaum, 2001a, pp. 414–20).

But the judgment on which Nussbaum dwells the most in her account of the moral education of compassion is the eudaimonistic judgment – for, as I said, it is the one in which we are most lacking due to our spontaneous tendency to ignore the conditions of those who are further away. According to Nussbaum, even in this case it is possible to fix this flaw through philosophical work and imagination. She proposes to gradually extend our attention to the fate of others, expanding the boundaries of individuals we are interested in. Through moral reflection and empathetic imagining of the lives of those more distant, we should seek to include others within the sphere of our goals and ends. (This is why, according to Nussbaum (1995, pp. 31–39), moral education needs not only ethical principles drawn from philosophical reflection, but also the support of literature: enabling us to see the world from the perspective of different people, but with whom we share a common humanity, it allows us to develop the capability of identification and a ‘generous construction of the seen’.)

To explain this process, Nussbaum (1994, pp. 342ff.; 2001a, pp. 388–92) borrows a metaphor from the Stoic Hierocles, who imagined people's sphere of interests as a series of concentric circles: the first one includes the single individual, then the closest affections, family members, neighbours, fellow citizens and so on, up to the whole universe. The task of ethical reflection is to progressively decrease the distance of these circles from the centre, making the effort to gradually include all human beings among those we care about. This form of ‘widened’ compassion can, according to Nussbaum, overcome the human tendency for parochialism and, through proper education, elicit a greater commitment to the whole of humanity.

5. Why Nussbaum's Case Begs the Question

Unfortunately, despite its refinement, this strategy has some problems. A number of authors (McQueen, 2014; Bloom, 2016) have raised doubts about the possibility of actually achieving such an extension of people's horizons. After all, it may seem unrealistic for one to develop a concern for all of humanity, especially considering our natural inclination to empathise more with individuals than with large groups of people (Small & Loewenstein, 2003; Kogut & Ritov, 2005; Slovic, 2007). But that is not the criticism I want to focus on. Rather, I mean to argue that Nussbaum's account of the development of moralized compassion conflicts with her idea of using this emotion as the motivational grounding of the capabilities approach. Having shown the beliefs underlying compassion in Nussbaum's own theory, it is possible to return to what I have called the case for moral compassion and understand where the contradiction lies.

In section 3 I sketched the argument in the following simplified form. The supporter of the capabilities approach knows that a life is flourishing if it is characterised by the free exercise of some central capabilities (characteristics of human existence), which define a life worthy of a human being. (S)he knows that, without these capabilities, one cannot live a good life – a life worthy of human dignity. Thus, faced with someone who is unfairly lacking in central capabilities, the supporter will feel compassion for them; and this emotion provides him or her with the motivation to help them. The picture can now be enriched thanks to the definition of compassion that we have analysed. As we have seen in the previous section, for Nussbaum it consists of the three judgments of size, nondesert and the eudaimonistic one. Seeing someone devoid of some central capabilities, the supporter of the approach cannot help having the first two beliefs: (s)he realises their plight; they are not just suffering a trivial damage but something which prevents them from true human fulfilment. Furthermore, (s)he will acknowledge that, often, the suffering is not a victims’ fault but the result of an undeserved condition, beyond their control. However, following Nussbaum's own account, in order to feel compassion, it is necessary to have also an eudaimonistic judgment, i.e., to consider in some way the victims’ well-being as important for one's own flourishing – or, as Nussbaum (2001a, p. 319) puts it, ‘as a significant part of [one's] own scheme of goals and ends’. This means that the capabilities approach assumes from the start that people are concerned for the happiness of others:

The capabilities approach is able to include benevolent sentiments from the start in its account of people's relation to their good. This is so because its political conception of the person includes the ideas of a fundamental sociability and of people's ends as including shared ends […]. Prominent among the moral sentiments of people so placed will be compassion, which I conceive as including the judgment that the good of others is an important part of one's own scheme of goals and ends. (Nussbaum, 2006, p. 91, my emphasis)

But, as Nussbaum (2006, p. 157) herself acknowledges, ‘[r]eal people often attend to the needs of others in a way that is narrow or arbitrarily uneven’. As we have seen, she is aware that our spontaneous compassion is narrower than it should be to motivate a wide altruism such as that required by the capabilities approach – which prescribes to overcome even national boundaries and barriers of species. As the reader will remember, Nussbaum proposes to overcome this problem through a moral education that teaches how to expand compassion beyond the narrow limits in which it usually dwells. However, this argument begs the question. For, according to Nussbaum's own account, this education to feel a wider compassion should be achieved through political institutions and a public culture inspired by the principles of the very capabilities approach, which however already assumes as a premise a broader moralized compassion than the one we spontaneously feel. As a result, the relation between appropriate emotions (especially compassion) and moral principles is circular: the former have a true cognitive basis when inspired by the latter; but the latter already presuppose the former to be present in actual people.

Not surprisingly, this circularity is akin to the one found in Aristotle's moral system between habit and practical wisdom. As it is well known, he argued that moral action is the result of the subject's good character, that is, of the habit of acting in a virtuous way. To acquire this habit, it is necessary to behave virtuously over and over again; and, to do this, one needs to be endowed with practical wisdom (phronesis). But at the same time, to be wise (viz., to have the knowledge of what the right action is in different circumstances), it is necessary to already have a virtuous habitus – that is, to already have the habit of behaving in a good way. Something similar happens, in Nussbaum's case, between the moral-political theory of the capabilities approach and the appropriate emotions. Unlike contractarianism, her account does not see justice as an agreement between individuals interested in maximising their own advantage, but assumes the presence of benevolent emotions that motivate people to guarantee everyone a minimum threshold of capabilities to flourish, regardless of the benefit they can derive from that. However, it is possible to develop such emotions only by embodying true moral beliefs, supporting (and displaying through institutions) the very theory of justice that they are supposed to support.

The petitio principii can easily go unnoticed, for Nussbaum does not make the case for moral compassion as explicit as I have done. Rather, she often appeals to vague reasonable statements suggesting a mutual (or dialectical) support between principles and emotions. For example, she claims that, since ‘benevolence can give indeterminate results’, ‘[w]e do not try to generate principles out of compassion alone, but, instead, we seek to support them and render them stable through the development of a compassion that is attuned to the political principles for which we have argued’ (Nussbaum, 2006, p. 91). But in other passages it emerges clearly that the capabilities approach ‘envisages human beings as cooperating out of a wide range of motives, including the love of justice itself, and prominently including a moralized compassion for those who have less than they need to lead decent and dignified lives’ (Nussbaum, 2006, pp. 156–57, my emphasis). Unfortunately, the moralized compassion presupposed by the capabilities approach can be developed, as I have shown in section 4.1, only after our natural compassion is extended through a reflection inspired by the same moral theory.

6. Emotions and Capabilities: Is It Possible Not to Beg the Question?

In the previous section, I showed how the relationship between appropriate emotions and moral principles, central to Nussbaum's theory, begs the question. Is there a way to solve this problem, saving the capability approach from the circularity? A first attempt, which can be indirectly derived from some of Nussbaum's arguments about education, is inspired by Aristotle. In his practical writings, he addressed the problem of the circularity between the habit of virtue and practical wisdom through the educational system. He imagined that the only way to acquire a good character was to develop it during one's childhood, following the prescriptions of parents and good teachers. After one is initially forced to act virtuously through public education, over time they will embody it more and more, developing the habitus necessary to acquire practical wisdom – that is, the ability to evaluate what is the right thing to do in different circumstances. This capability will then be perfected by the subject in the course of their life, confronting each time with different situations (Aristotle, 2002, Nicomachean Ethics, V.2, X.9; cf. also Nussbaum, 1994, pp. 96–101).

In a somewhat analogous way, it is possible to address the objection I raised in section 5 by referring to Nussbaum's (2006, pp. 408–15) concept of education, understood as an incremental process, i.e., aimed at gradual improvement. She argues that, through a proper education at school and through good political institutions, the new generations will be somewhat more equipped than the previous ones to deal with the new ethical challenges, showing a greater interest in the well-being and rights of subjects of justice previously overlooked. Over time, these new generations, trained in having a wider concern and compassion than their parents, will modify public culture. As a result, they will produce an educational system capable in turn of expanding the horizons of the following generations even more, in a constant approximation to the ideals of justice.

This process of progressive widening of our concern for others has some merits. First, it is a rather realistic way of realising Hierocles’ metaphor of expanding the circle of our interests. Moreover, it seems to have some historical precedents: in many respects, this is what happened in liberal democracies, where over time we gradually enlarged the number of people whose lives are considered of equal value and worthy of protection – roughly, this was the path for the abolition of slavery, or the recognition of equal rights for women, blacks, gays and lesbians, etc. However, while someone may find it appealing, even this more sophisticated version of the case for moral compassion fails to really overcome the theoretical problem I have raised. In fact, even this version of the argument takes for granted that the goal of education is to widen the sphere of concerns of the new generations, including more distant people who were previously not considered. But such a goal is warranted only insofar as we already accept Nussbaum's ethical theory, which – as we have seen – in turn assumes that current subjects possess a broad moralized compassion. Hence, even appealing to the role of education as an incremental process fails to fully solve the puzzle I raised.

At this point, a supporter of the capabilities approach might object that people's emotions are not a problem worth considering for Nussbaum's theory. After all, one could say, a theory of justice aims at defining what is right to do, regardless of whether people emotionally accept it. However, this would be a mistake. First, because, as I have shown in section 3, Nussbaum repeatedly emphasises that one of the key features of her theory is precisely the role she acknowledges to benevolent emotions, in contrast to traditional contractarianism. Second, throughout her career, she has always been sensitive to the so-called ‘problem of motivation’ (i.e., what leads us to act morally once we know what this implies). She takes up Rawls's claim that the well-ordered society needs a ‘reasonable moral psychology’ in both Political Emotions (Nussbaum, 2013, p. 9) and Upheavals of Thought (Nussbaum, 2001a, p. 402) – where she speaks of ‘reasonable political psychology’. Introducing her analysis of the role of compassion in public life, she writes:

Why should such a conception deal with emotions at all, it might be asked? The answer is, plainly, that any political conception needs to concern itself with citizens’ motivations, both in order to ensure that the conception is feasible in the first place – does not impose impossible strains on human psychology – and also in order to ensure that it has a decent chance of being stable over time. (Nussbaum, 2001a, pp. 401–2)

Given how much Nussbaum emphasises the role of compassion in distinguishing her capability theory from contractarianism, and her insistence on the significance of emotions in motivating citizens of a well-ordered society, ignoring the problem I have raised would be contrary to Nussbaum's own project. Instead of overlooking the role of emotions for her account, I would like to conclude by trying to sketch another way to deal with this challenge, which seems to me more promising in saving the role of emotions as a basis for the capabilities approach. I suggest borrowing the distinction, proposed by Ronald Dworkin (2000, 2011) and accepted by other authors (Appiah, 2005), between morality and ethics. In his jargon (Dworkin, 2000, p. 485n1), morality is the realm of our duties towards others (i.e., it ‘includes principles about how a person should treat other people’), while ethics has to do with what we owe ourselves, ‘includ[ing] convictions about which kinds of lives are good or bad for a person to lead’. (Also this distinction is rather Aristotelian, for the crucial question of the Nicomachean Ethics is not, as for the moderns and contemporaries, how we should behave towards others, but above all what it is to live one's life well.) According to Dworkin (2011, pp. 191–218, 255–64), each of us aspires (and should aspire) to give our lives a value that is not simply subjective (that is, linked to particular preferences, tastes, opinions we happen to have) but objective, making one's existence ‘a successful performance rather than a wasted opportunity’ (ibid., p. 203). Such a goal parallels Aristotelian eudaimonia, or happiness, as full self-fulfilment. However, he argues, it is possible for us to achieve such an objective value in our own life only insofar as we recognise all human lives having the same objectivity. For it would be very difficult to justify that one's life has an objective value – i.e., a value independent of our personal preferences, predispositions, talents, and contingent desires – claiming at the same time it is not equally important that the lives of other human beings achieve this same goal. Such a position would be inconsistent. In this way, starting from the idea (ethics) that it is objectively crucial for each of us to strive to make our life successful, we warrant the motivation to behave morally – viz., to recognise that others also have the same crucial responsibility to give their lives an objective value. This is what it means, for Dworkin, to live with dignity and respecting the dignity of others (i.e., to respect their efforts to give their existence an objective value, according to one's own ideals).

The same line of argument can be applied to the capabilities approach. Following Dworkin's insights, we could find the emotional support for that moral theory not in compassion but in other affective sources, such as the desire to have a good life, endowed with objective value.¹³ The path would be the following: as human beings, the supporters of the capabilities approach aim to achieve a good life (or eudaimonia), which has an objective value. They know that it can be a successful performance only through the free exercise of one's capabilities in ways and fields one considers valuable. However, it is possible to claim the objective importance of the flourishing of one's life only if one coherently recognises equal importance also to the fulfilment of the life of each individual. This does not mean that we should be as committed to the fulfilment of other people's lives as we are to our own. On the contrary, everyone has a special responsibility to make one's life a successful performance. But, if we are to be consistent, we must seek to fulfil this ethical commitment while equally respecting other people efforts to make their existence valuable. As a result, the supporter will be motivated to set a political system that shows equal consideration for each individual's efforts to achieve a good life, in ways that they value. This is equivalent, from the point of view of the capability theory, to guarantee a minimum threshold of capabilities to have the opportunity to lead a flourishing life.

The account I have sketched is not complete. It does not deal with some crucial problems of the capabilities approach (and many other ethical-political theories). For example, I have not addressed the issue of what the minimum threshold of capabilities concretely is; nor have I given clues on what to do when two capabilities conflict, or on to what extent we are justified in demanding sacrifices from those who are living a good life to help people who do not have the same chance to succeed because of conditions of injustice. My purpose in this article was more theoretical: to analyse the role of emotions in grounding Nussbaum's capabilities approach. I have shown the importance, not yet considered enough, of affects as a motivation for moral behaviour in Nussbaum's account. Through my analysis, I hope to have underlined some problems in her proposal to take compassion as the basis of morality, and to have suggested an alternative way – but still consistent with her approach – of addressing the issue of motivation by appealing to other affective sources.