Claim (1): Love is a matter of justice

Here I am concerned with love in general, rather than with any particular kind such as romantic, companionate, parental or filial love. I often use illustrations drawn from parental love, which are best suited to make the case that the extent to which individuals enjoy love is a matter of justice. Yet, the scope of the paper goes beyond parental love. Love as understood here is personal, which means that it is directed at, and responsive to, the individuality of the beloved rather than being directed at, and responsive to, what particular individuals have in common with wider groups of individuals, as is the case with Christian love, Buddhist love, or love for humanity. Being personal in this sense, love is also partial: we love some individuals and not others. On most interpretations, personal love involves responding to the individuality of the beloved with a robust disposition to care for or be attached to that individual. But the present analysis is equally relevant to the alternative view that love consist in an ability to pay special attention to the beloved. I also assume that, precisely because it is directed at the individuality of the beloved, the attitudinal dimension of love involves an element of spontaneity which is constitutive to how love is valuable for the beloved.

Two arguments support the conclusion that being loved and, perhaps, being able to love, are matters of justice. If at least one of them is successful, claim (1) can be established. The first argument assumes luck egalitarianism:

Argument 1.1.

1. (a) It is unfair, and thereby unjust, for some to be worse off than others through no fault or choice of their own (luck egalitarianism).

2. (b) Being loved significantly contributes to how well a person’s life goes.

3. (c) Love is, to a significant extent, a matter of brute luck.

Therefore

1. (1) Love has to be taken into account when assessing the justice of a distribution, and, in this sense, it is a matter of justice.

Although popular amongst contemporary philosophers, luck egalitarianism is not endorsed by everybody. A second argument to the same conclusion assumes that justice requires only that which makes lives go sufficiently well; its more modest upshot is that being loved and able to love sufficiently are matters of justice:

Argument 1.2.

1. (d) Justice requires that people have what they need in order to lead sufficiently good lives (sufficientarianism).

2. (e) People need to be loved and to be able to love at least sometimes and to some degree in order to lead sufficiently good lives.

Therefore

1. (1) Love is a matter of justice.

Argument 1.1.

1. (a) It is unfair, and thereby unjust, for some to be worse off than others through no fault or choice of their own (luck egalitarianism).

In Cohen’s (Reference Cohen2009, 7) words, ‘an unequal distribution whose inequality cannot be vindicated by some choice or fault or desert on the part of (some of) the relevant affected agents is unfair, and therefore, pro tanto, unjust’. This is an influential conception of justice and its fundamental intuition about fairness is shared by philosophers who cannot be properly described as luck egalitarians. It is for instance present in Rawls’ understanding of justice, at least to the extent to which the moral goal of a theory of justice is to come up with a conception ‘that nullifies the accidents of natural endowment and the contingencies of social circumstance as counters in the quest for political and economic advantage’ (Rawls Reference Rawls1972, 5).

1. (b) Being loved significantly contributes to how well a person’s life goes.

‘Nobody has ever loved me’ is one of the gravest complaints that one can make about their life. Similar – or perhaps worse – is ‘I am not able to love anybody’.Footnote ³ Moreover, it is widely believed that love cannot be substituted by the enjoyment of other goods. Enjoying some kind of love to some degree seems to be a significant constitutive part of a person’s subjective and objective well-being.

This belief is supported by anecdotal evidence and testimonies of people who identify loving family and friendship relationships as the, or at least one of the, most important goods contributing non-instrumentally to their well-being; this led some philosophers to argue that in a just society people would enjoy loving relationships equally (Lynch et al. Reference Lynch, Baker, Lyons, Baker and Lyons2009). Love is also instrumentally valuable to creating agency: a vast feminist and psychological scholarship defends the importance of early loving relationships for becoming properly socialised, and moral, individuals (Ruddick Reference Ruddick1989; Held Reference Held1993; Hollway Reference Hollway2006) – characteristics which, in turn, are necessary conditions for leading a good life.

If it is true that being loved, especially early in life, contributes to the ability to value oneself as a unique individual, to become aware that one has intrinsic value and to acquire basic inner resources such as confidence and resilience, then enjoying loving relationships is what philosophers Wolff and de-Shalit (Reference Wolff and de-Shalit2007) call a ‘fertile functioning’ – that is, a desirable state that increases one’s likelihood to achieve additional desirable states.

Different luck egalitarians identify the metric of justice as either welfare, or capabilities or resources. The non-instrumental value of love is plausibly important enough to be necessary for a fully flourishing life – and hence love is directly relevant to theories that identify welfare, or opportunities for welfare, or capabilities, as the metric of justice. Whether or not one’s need for love is met also shapes that person’s opportunities for other goods; because love also has instrumental value, it can be plausibly seen as a kind of resource and therefore relevant to resourcist conceptions about the metric of justice. Love cannot, however, count as part of the metric of justice on a monistic resourcist view that identifies money as the relevant resource.

1. (c) Love is to a significant extent a matter of brute luck.

Perhaps, as adults, we can be held responsible to a great degree for the good as well as for the bad that love brings us to the extent to which we can choose the people with whom we relate. Yet, it is obvious that much of the circumstances of our relationships are not under our control – such as the conditions in which we meet those we love or lose them because they die. More importantly, to many of us it is somewhat mysterious what elicits our affection, desire, tenderness, longing, curiosity and attention. It is far from obvious that one can decide to love certain individuals and not others, even if the latter would make, according to one’s own judgement, perfect recipients of one’s love (Rainbolt Reference Rainbolt2016); nor is it possible to control who loves us. Further, we have little control over continuing to love another person or securing their continued love.

The claim that love is a matter of brute luck is most difficult to dispute if we think about children and parents: children do not have any choice of who their parents are and how well their parents love them.

From (a), (b) and (c) it follows that: (1) Love is a matter of justice.

Many are sceptical of luck egalitarianism; its critics worry that it is impossible to implement its ideal of fairness and some point out that even trying to implement it would have unacceptable consequences for personal relationships (Lanzenby Reference Lanzenby2010). They conclude that luck egalitarianism is best seen as a theory which, on its own, has little action-guiding potential. This conclusion does not undermine the point of this paper. It is consistent with believing that, if we could shape the world such that nobody’s life went worse than others’ for lack of love out of no fault or choice of their own, it would be a requirement of justice that we do it.

More radical critics believe that luck egalitarianism is implausibly describing as unfair, and pro tanto as unjust, too much of what we usually attribute to unavoidable unfortunate circumstances – and, indeed, the way in which love upsets fairness is likely to raise this objection. A widely endorsed alternative to luck egalitarianism is the view according to which the aim of justice is to ensure that everybody has enough to lead a sufficiently good life. In particular, according to the defenders of an increasingly influential view, an adequate understanding of justice should instead focus on the relationship between citizens, aiming to eliminate obvious forms of injustice such as exploitation, domination, oppression, exclusion and marginalisation (Young Reference Young1988; Anderson Reference Anderson1999; Scheffler Reference Scheffler2003). Many of these critics, who call themselves relational egalitarians, think that the real aim of justice is to ensure that everybody is sufficiently functional and socially included. Even if the aim of justice is to establish egalitarian relationships rather than some kind of distributive equality, just societies will not allow their members to fall below a certain threshold of well-being. I argue that various types of sufficientarians all have reason to endorse the second argument to the conclusion that love should be regarded as a matter of justice:

Argument 1.2.

1. (d) justice requires that people have what they need in order to lead sufficiently good lives (sufficientarianism).

Sufficientarianism is a modest standard of justice. However, those who reject pattered conceptions of justice, such as libertarians, deny that even minimal well-being is due to people as a matter of justice. For them, the overall argument of this paper is a reductio of the particular conceptions of justice to which I appeal.

1. (e) people need to be loved at least sometimes and to some degree in order to lead sufficiently good lives.

Needs of the non-instrumental, morally significant, kind are best understood in relation to harm. According to Wiggins’ (Reference Wiggins1998, 10) definition of intrinsic needs, I need X ‘if it is necessary, things being what they actually are, that if I avoid being harmed then I have X’.Footnote ⁴ Another feature of intrinsic needs is that they are not chosen, and that individuals cannot simply decide not to need what they need.

It then seems plausible that all people need to be loved: much time, energy and other resources are spent in the quest for love and, when love is found it never seems to appear futile (unlike other goals that we pursue relentlessly.) According to people’s own testimonies, lives devoid of any love or hope for future love are often very miserable for this reason. And it is not up to people to decide that they can do without love (although it can be up to an individual to decide they can do without a loving relationship with a particular individual.) This is a stronger claim than (b) while still weaker than the Beatles’ slogan.

The most charitable interpretation of Beatles’ slogan – which will not convince everybody – is that we have a very important need for love and that love cannot be substituted. Then, love’s absence entails significant harm and lives can hardly go sufficiently well if entirely deprived of love. This is vividly illustrated in the case of parental love: even those who are sceptical that adults need love in order to have sufficiently good lives will likely accept that children’s lives cannot go well enough in the absence of love from an adult. For this reason, Liao (Reference Liao2006, Reference Liao2015) and Richards (Reference Richards2010) have recently defended the controversial position that parents have a duty of justice to love their children and children have a correlative right to be loved. Both argue that parents are responsible to ensure that their children reach an adequate level of well-being and that their children’s developmental needs are met, and rely on empirical literature to show that none of these goals can be achieved in the absence of parental love. According to Liao: ‘Studies of children in institutions found, for example, that children who did not receive love but only adequate care became ill more frequently; their learning capacities deteriorated significantly; they became decreasingly interested in their environment; they failed to thrive physically by failing to gain weight or height or both; they suffered insomnia; they were constantly depressed; and they eventually developed severe learning disabilities’ (Reference Liao2006, 423).

Further, some philosophers and psychoanalysts writing on recognition draw on a Hegalian tradition to claim that love, in particular parental love, plays a central role in the formation of agency – in shaping personhood and autonomy (Benjamin Reference Benjamin1988; Honneth Reference Honneth1992). For example, it has been claimed that love is necessary for self-trust (Anderson and Honneth Reference Anderson, Honneth, Christman and Anderson2005). If having experienced love is crucial for autonomy and personhood – a claim also endorsed by the feminist ethics of care, at least when applied to parental love (Ruddick Reference Ruddick1989; Held Reference Held1993) – then its absence can easily undermine our standing as political equals. Therefore, relational egalitarians as well as other sufficientarians have reason to be concerned with its distribution.

As Liao’s critics noted (Cowden Reference Cowden2012), empirical studies only show that good emotional relationships between children and their caregivers are essential for children’s flourishing and that in their absence children’s development is seriously impaired. But we do not know which aspect of these relationships is crucial for flourishing. As far as the empirical studies are concerned, one may conclude that children’s flourishing is entirely due to the behavioural aspects of parental love, such as expressions of affection towards their children, attentiveness to their individual needs, quality time spent together and parents’ stance of partiality towards their children. How much does it matter to children that such behaviour is motivated by their parents’ partial response to their children’s individuality? It may be very difficult to devise empirical investigations to establish this. First, the emotions of love are very difficult to capture, define and measure. Second, it is very unlikely to encounter sufficient cases of the above-mentioned behaviour unsustained, in the long run, by genuine love, that is by responsiveness to the individuality of the child. While I do not think that the empirical evidence gives (or, possibly, could give) conclusive support to the claim that responsiveness to the individuality of the child is a significant part of what makes love so important for children, I think we have nevertheless good reason to give credence to this claim. The good of personal love is not reducible to the good of beneficial actions. Because when done out of love these actions are motivated by the individuality of the beloved, love is an affirmation of the value of the beloved – a point to which I return below. And, plausibly, it is love’s affirmation of the beloved’s value that gives love an important role in the formation of autonomy and personhood.Footnote ⁵

From (d) and (e) it follows that, especially during the formative years of childhood: (1) Love is a matter of justice.

The sufficientarian argument makes no reference to personal responsibility or individual choice. On the one hand, obliterating the issue of personal responsibility makes the sufficientarian argument a generous frame for understanding the relationship between love and justice: if somebody’s life is entirely devoid of love due to that person’s choices, and as a result an intrinsic need of that person is unmet, this situation would still count as unjust according to the sufficientarian account.Footnote ⁶ Unsurprisingly, one of the most influential sufficientarian theories of justice, which identifies capabilities as the metric of justice, explicitly lists a capacity to love and be loved amongst the relevant capabilities (Nussbaum Reference Nussbaum2006). Drawing on psychological literature, Liao (Reference Liao2015) argues that the acquisition of these capacities requires that one has experienced parental love.

On the other hand, the second argument is less demanding, as it need not describe as unfair or unjust all inequalities in how much and how well we are loved (and for which we are not responsible). If luck egalitarianism is true, then all inequalities in love that are not due to our own responsible choices are unfair unless they are compensated for. In practice, this would certainly amount to detection of massive injustice in the world. On most sufficientarian accounts there is no injustice in the fact that we are unequally loved as long as we all receive enough love for our lives to go well.

To illustrate the different implications of arguments one and two, the first deems a situation where parents love some of their children more as unjust while the second deems it just as long as all children receive sufficient love.

Claim (2): There cannot be a duty to love

Kant formulated an influential endorsement of (2): ‘Love is a matter of feeling, not of willing, and I cannot love because I will to, still less because I ought to (I cannot be constrained to love); so a duty to love is an absurdity’ Kant (Reference Kant and Gregor1996, 161).Footnote ⁷ Recently, however, several philosophers have argued that children have a right to be loved and that parents have a duty to love their children (Liao Reference Liao2006, Reference Liao2015; Richards Reference Richards2010; Ferracioli Reference Ferracioli2014). Liao, who defended a right to be loved, and Luara Ferracioli, who defended a duty to love – both driven by the importance that love has for children’s well-being and development – usually read the Kantian position as being about what people can and cannot do – so, ultimately, as grounded in an empirical claim (which they reject). The same interpretation of Kant’s position is implicit in Liao’s critics and other philosophers who draw on Kant in order to deny that there can be rights and duties to love (Brake Reference Brake2011; Cowden Reference Cowden2012). I will therefore first discuss:

Argument 2.1.

1. (f) We cannot fully control whether or not we love.

2. (g) ‘Ought’ implies ‘can’.

Therefore:

1. (2) There cannot be a duty to love.

As I will explain, it is difficult to deny (f). But even if (f) was false, there is a philosophically more interesting, and overall more plausible way of understanding Kant’s position as being not about the limits of individuals’ abilities – and hence about an empirical matter – but about what counts as ‘love’ – and hence about a conceptual matter. On this second reading, if we could fully self-induce all the dispositions and emotions that we typically identify as ‘love’ we would nevertheless fail to love, because, by assumption, love must be at least in part caused by something outside the lover – by the beloved. (Kant’s formulation, at least in translation, appeals to human ability, yet concludes that a duty to love is absurd rather than impossible. It therefore remains ambiguous between an empirical and a conceptual interpretation.)

The next argument draws on Philip Pettit’s recent account of love as a robustly demanding good to explain why love-like behaviour that does not respond to the individuality of the beloved is conceptually different from love. A conceptual argument about the impossibility of a duty to love, however, would be of limited interest if it was merely semantic. I compare genuine love to love-like behaviour that fails to be adequately motivated by the individuality of the beloved and claim that the intrinsic value of the former comes from the fact that it is given out of inclination and with a measure of spontaneity. Therefore, we have good reason to employ a concept of love that makes a duty to love absurd. One may, and possibly should, cultivate certain dispositions – including dispositions to notice, appreciate and become responsive to others – that make genuine love more likely. But the intrinsic value of love is realised only when inclination, and hence spontaneity, play a role in loving. This is:

Argument 2.2.

1. (h) People need some affection, care or attention given spontaneously, out of inclination.

2. (i) Therefore the affection, care or attention resulting exclusively from one’s self-controlled behaviour would not meet the specific need for love.

3. (j) Therefore meeting the need for love cannot be a matter of duty.

Therefore:

1. (2) There cannot be a duty to love.

I now move on to defend these arguments.

Argument 2.1.

1. (f) We cannot fully control whether or not we love.

As noted above, love cannot be entirely a matter of volition: we cannot bring ourselves to love other people just because we want to. This is a familiar claim, put forward by philosophers and non-philosophers alike. From Kant to Jane Austen, there is a consensus that genuine attachment cannot be fully commanded. Since love is responsive to the particularity of the beloved, it involves the individual taste of the lover and her selective sensitivity and attention – features that we can cultivate but not entirely control. Suppose Jane’s love for Jill may be renewed every time Jill displays her quirky, shy or tender sense of humour; the fact that she notices and appreciates this feature in Jill and not in others (at least not in a love-conducive way) is not entirely up to Jane. Most of us believe we are not able to engineer the partiality that defines personal love.

While the more familiar illustrations of (f) come from romantic or companionate love between adults, the love of parents for their children can also illustrate this point. Parents who, in spite of their best efforts, cannot love a particular child (for instance, because they are themselves struggling with depression), or cannot love all of their children equally, are not unheard of. This last example shows why (f) represents a problem for realising fairness with respect to love on a luck egalitarian understanding of justice. If you are less loved than your siblings through no fault of your own, and if being less loved than your siblings has a negative impact on your opportunities, then unequal parental love entails unfairness. And if love is important for justice because we have an intrinsic need for enough love, then parents’ inability to love a child enough will create injustice even according to the more modest principle of sufficiency.

The idea that there cannot be a duty to love has been contested, with respect to both parental and romantic love. Defenders of a duty to love children ground their argument in the importance of love for the beloved’s well-being to argue that love is owed to children as a matter of justice (Liao Reference Liao2006, Reference Liao2015; Richards Reference Richards2010). They also argue that love is, to some extent, under our control because we can cultivate loving dispositions towards individual people. Iddo Landau and Matthew Liao illustrated ways in which individuals can take action that is very likely to evoke, maintain or renew marital and parental love (Landau Reference Landau2004; Liao Reference Liao2006). These arguments are persuasive but limited. As their critics show, individuals can achieve only that much success in their attempts to self-induce affection towards or caring for, a particular individual (Brake Reference Brake2011; Cowden Reference Cowden2012; Rainbolt Reference Rainbolt2016). Affection and caring (as well as special attention) respond to some of the qualities of the beloved; this claim need not presuppose the controversial view that we love people because of their qualities, but the more modest one that sufficient change in certain qualities of the beloved can challenge, and sometimes destroy, love (Brake Reference Brake2011).

The claim that love cannot be commanded can be read as an empirical truth. But what if we were able to overcome the typical limitations of controlling love by making love impervious to changes in the beloved – what if we were able to create ‘the love drug’? A number of philosophers have recently defended the use of neuro-enhancements to improve love (Nyholm Reference Nyholm2015). Even if the love drug was available, however, its advocates suggest its use as a consolidator of already established love. None defends the claim that we could or should make available a love drug that makes it possible to assign, to each individual in need of love, someone to provide the affection, care or attention they could not provide in the absence of the drug. Nor does anyone suggest that love drugs could or should be used to ensure that everybody enjoys love equally. Defenders of the claim that we should make available a love drug propose it as a solution for expanding the scope of control we have over love, rather than bring it entirely under our control. As I show below, this is wise.

1. (g) Ought implies can – one cannot be morally required, even less be under a duty to, do something one cannot do.

The most uncontroversial interpretation of this fundamental, widely endorsed principle, is that an agent is not violating any duty by not doing what she cannot do – no matter how much value that action would realise, should she able to do it.

One cannot logically exclude the possibility of a world in which, serendipitously, everybody is sufficiently or equally loved. And there might be many things we can do in order to move closer to this world by giving people the time and leisure to be with and focus on others, educating them to cultivate loving dispositions and relationships, containing mental and emotional disorders and mitigating their effects. However, if the present argument is correct, our intentional action and hence duty will have to stop at doing whatever it takes for love to flourish for all – people to spontaneously respond with love to some others. Yet, all this amounts to is a (collectively held) duty to enable love.

From (f) and (g) it follows that: (2) There cannot be a duty to love.

Argument 2.2.

The worry concerning drug-induced affection, care or attention is that, being drug-induced, they would not be responsive to the individuality of the beloved. This amounts to a conceptual claim: that love over the creation of which the agent has full control – and which therefore could be given out of duty – is not really the partial, personal love that is so important to our well-being. If who you are plays no role in explaining my love for you, I may feel for you – say – humanitarian, or Christian, love but not personal love.

Everyday uses of ‘love’ tend to involve, conceptually, that love cannot be given at will. The non-volitional element of love explains why we believe that attempts to buy or sell love, or to extract it by force are not merely doomed to failure, but confused: If I were to report success in my attempt to fly without the help of a machine you would think that I am deluded or lying. But if I were to report success in buying someone’s love you should suspect that I fail to grasp the meaning of ‘love’. The motivation for loving someone – the process through which we get to respond with love to a particular individual – is part of what defines the complex of dispositions and emotions that we call ‘love’. So, for instance, if it turns out that we have been badly mistaken in our beliefs about a person we love, or that we have been drugged into feeling affection for her, we would think we don’t actually love her even if our subjective experience is identical to that of genuine loving. In the case of being drugged we may think that we are in the grip of an obsession rather than experiencing the genuine attraction of love.

Philosophers usually follow the common-sense use of ‘love’ in this respect. To take a recent example, Pettit (Reference Pettit2015) defends an account of love as a good which is robustly demanding because it involves a stable disposition to care for the beloved (across a certain range of possible world) and also because this disposition must be particularised or relational, that is triggered by the individuality of the beloved. Unlike dutifulness, which an agent should be able to trigger independently from who the particular identity of those to whom she owes duties, the dispositions that are constitutive of the love for a particular individual can only be triggered by that particular individual. Or else, it fails to count as personal love.

So (2) is best understood as a conceptual claim: for affection, care or attention to count as a genuine expression love, they should be responsive to the individuality of the beloved, and hence, at least to some extent, outside the lovers’ control. One may insist that this is a semantic, rather than a normative, point. Independent of what we call ‘love’, the relevant question seems to be whether we could achieve a fair distribution of what we find valuable about love, by better training our emotions and behaviour or by using a ‘love drug’ that induces love-like behaviour. If we had the means to control love-like behaviour, would a duty to do so achieve a just distribution of all that is valuable in love? I believe not, and here is an argument to that conclusion:

1. (h) People need some affection, care or attention given spontaneously, out of inclination.

What is the value of personal love? There are situations in which it seems, intuitively, that our yearning for personal love is so great that we cannot but feel disappointed when, instead of it, we receive mere dutiful love-like behaviour – even when the latter is as such praise-worthy. Consider Michael Stocker’s example of the hospital visit:

This story illustrates an important need that personal love fulfils: that some people display partialityFootnote ⁸ towards us – perhaps what Kant has called pathological inclination. To enjoy the partiality of a lover is part and parcel of the good of being loved by that person. Moral philosophers are usually interested in how partiality hinders our ability to comply with moral demands, which makes partiality regrettable. However, to the extent to which partiality is constitutive of personal love it also represents an important good. Being sometimes on the receiving end of loving partiality seems crucial for feeling valued as an individual. Some think that personal, and thereby partial, love is sufficiently important to override reasons of duty in some cases: Williams (Reference Williams1981, 18) described the person who wonders whether it is morally acceptable to save the life of his wife, rather than that of a stranger, as having ‘one thought too many’. Now, in cases where rescuers do not have one thought too many, their beloved may have two reasons to rejoice: that they were rescued, but also that they are so valuable to their rescuer that considerations of duty did not enter the rescuer’s deliberation.

Going back to the example of children’s need to be loved by their main care-givers, it is indeed easy to make sense of a child complaining that the care, affection, attention and time she receives from her parents are given entirely out of duty. This child has reason to feel that something very important to her is missing. She might feel, for instance, that she is treated impersonally – that the affection she receives has nothing to do with her individuality. More generally, when people display love-like behaviour towards us, but we suspect that they are acting out of mere duty, we may feel valued as human beings, but certainly not as particular individuals, and we have reason to regret the later. Personal love seems unlike any other good the distribution of which can be a matter of justice. When individuals receive due amounts of, say, food by agents whose behaviour is entirely motivated by duty, there is no complaint left with regard to food distribution. But when they are at the receiving end of affectionate, caring, attentive behaviour entirely motivated by duty there remains a complaint: these individuals do not enjoy the full good of love.

What is the good of personal love such that one fails to provide it if one’s love-like behaviour fails to be, at least in part,Footnote ⁹ motivated by the reality of the beloved? It is plausible that personal love is so important to the beloved because it is an affirmation of her value. Some believe that love is an appraisal of the value of the beloved (Velleman Reference Velleman1999), others that it is a bestowal of value on the beloved (Singer Reference Singer1994). In either case, in love, aetiology as well as phenomenology count. Love-like behaviour that fails to be caused by the beloved cannot deliver this good. As Nyholm (Reference Nyholm2015, 197) puts it: ‘Love is … a sort of confirmation that we are, as we might put it, “lovable” in the sense of being able to inspire or call forth such dispositions in another (namely, the lover). Part, but not the whole, of what our lover gives us in loving us, in other words, is this confirmation of us as really being able, or having the power, to inspire loving devotion in another’.

Personal love can give such confirmation because, in being loved by someone, the beloved directly contributes to her lover’s well-being. In order to fully satisfy a person’s interest in being loved it is necessary that the lover’s own interest is, to some extent, involved in the loving relationship. Think, for instance, of a parent playing with her child: because she loves the child, the parent takes particular pleasure in watching the child have fun and this – the parent’s genuine delight – is part of what satisfies the child’s need for love. Or think about what it means to grieve for a loved one who died, as opposed to regretting the death of someone about whom you merely cared. In the latter case you may be sorry for all the good things that person will be missing out on, but not for yourself. By contrast, grieving for someone you love means to also perceive their death as a loss to yourself – to regret not only that the deceased cannot, say, laugh with and talk to other people anymore, but also that you cannot laugh with them and talk to them anymore.

If one accepts step (h), one will accept that:

1. (i) Therefore the affection, care or attention resulting exclusively from one’s self-controlled behaviour would not meet the specific need for love.

It follows that:

1. (j) meeting the need for love cannot be a matter of duty.

Therefore:

1. (2) There cannot be a duty to love.

This last argument supplements the conceptual point with an axiological one: it states that ‘love’ given entirely out of duty would not be love and that we have good reason to hold on to this concept of love because we need affection, care or attention that spring from our lovers’ partiality towards us; this requires some degree of spontaneous inclination towards us. If our emotional needs were more modest – if we could do with affection, care or attention of an impersonal, impartial kind, then, perhaps, a duty to meet this need could exist.

Note that, if the argument of this section is correct, the familiar danger of conflating issues of feasibility with issues of desirability is not present in the case of love. A crucial test for deciding whether something is a matter of justice is to ask: ‘If we could bring that thing about, would we have a duty of justice to do it?’ For instance, on purely luck egalitarian grounds, if we could, at no costs, shuffle eyes between the people with perfect sight and the blind, such that everybody ends up with one functional eye, we should do it. The test cannot be run with love if it is conceptually true that there cannot be a duty to love. If (2) is true, whatever one thinks about the relationship between feasibility and justice, one will want to reject the claim (1) that love is a matter of justice if one subscribes to claim (3) that all claims of justice are conceptually connected to claims of duty. It is to this last claim that I am now turning.

Claim (3): All claims of justice are conceptually connected to claims of duty

The third claim is almost universally endorsed. The most straightforward way of interpreting it is to say that something can be matter of justice only if it is some agent’s duty to bring that thing about. This interpretation would make claim (3) a good candidate for rejection, and rejecting claim (3) would dissolve the paradox.

It seems false that something cannot be a matter of justice if nobody has a duty to do something about it: sometimes there is no agent whose duty it is to prevent or rectify a particular injustice, without this being a reason to think that the state of affairs in case is just. Examples include past injustices whose victims and perpetrators are dead such that no rectification is possible, nor is there any agent who may have a duty to compensate. (Imagine, in addition, that no living individual either benefits or suffers from the past injustice.) Assume there have been cases of one group enslaving or destroying another group that happened so far back in the past that no living person’s life is affected by it. In these cases some duties existed – and have been violated – in the past. One may speak, here, of past duties.

There is a second type of case when something can be a matter of justice – at least according to an account of justice that is insensitive to feasibility constraints – and in which there are no duties, present or past. Think for instance of natural catastrophes that cannot be prevented or foreseen, and that cause irreversible damage; or, similarly, of people dying very young from unpreventable and incurable diseases. Since the victims are dead, there are no duties of compensation. In all these situations some people’s lives went (much!) worse than others’, through no choice or fault of their own. In these cases, however, there exist what we might call counterfactual duties: if an agent could have done something for the victims that agent ought, as a matter of justice, to have done it.

Therefore, the more precise interpretation of (3) is: ‘For something to be a matter of justice there should be a conceivable duty to bring that thing about’. That is, matters of justice are conceptually connected to duties: present, past or counterfactual.

Yet, unless (2) is false, in the case of love there are no duties – present, past or counterfactual. The predicament of the person who needs love, and whose life goes worse than others’ because she lacks love is similar to that of people caught in a natural catastrophe which could not be averted in that it is not the result of human action and could not be avoided by human action. But here the similarity stops. If one could save the victim of a natural catastrophe, one ought to do it; by contrast, no change in the feasibility constraints that agents may face could generate a duty to love.

This is not to deny that, sometimes, a failed duty is part of the causal history leading to the lack of love. One illustration of this claim is the situation of large numbers of children who grew up without parents in the 1980’s and 1990’s on Romanian streets. This situation was directly caused by policies banning contraception and abortion, and by the Romanian state’s failure to live up to its duty to provide adequate care for abandoned children. Presumably, in the absence of these policies and failures of justice some of the children deprived of parental love would not have been born, and many of those who have been born would have enjoyed adequate parental love. But certainly not all maldistributions of love flow from past failed duties: there is no reason to think that, had the Romanian state behaved justly towards all its children, they would have all enjoyed equal, or even enough, love.

A paradox?

Unless one of the three claims can be shown to be false, a very unsettling conclusion follows: not merely that perfect justice is not feasible – something few would dispute – but that we cannot think about love and (at least some kinds of distributive) justice together in a consistent way. Which of the three claims should then be rejected?

I believe we should not abandon the major premises supporting (1) easily. In particular, abandoning luck egalitarianism would entail accepting that sometimes unfairness is irrelevant to considerations of justice, and this is a high theoretical and possibly practicalFootnote ¹⁰ price. A better way to resist (1) would be to show that love as defined here is an important, but not unique, form of validation of our individuality. This is a theoretical possibility, but the burden of proof seems to be on whoever thinks that something other than personal love can deliver what I identified here as its non-instrumental good.

I also believe, for reasons discussed at length in the second section, that we should continue to endorse (2). (2) Could be rejected by showing that the good of love whose distribution is a matter of justice can be realised even when love-like behaviour is not motivated by the beloved’s individuality. This way out of the paradox is to deny that we actually need the kind of affirmation of our value that personal love gives us. Perhaps the importance we attach to our value as unique individuals is misguided. One may think that individuality itself is overrated and that human beings would be better off if, instead of love, they should seek impersonal and impartial affection (such as, for instance, Buddhist love or a secular version of Christian love). This possibility goes very much against the grain of our culture, which is deeply attached to the value of individuality, but possibly more coherent with alternative ways of valuing and therefore not impossible to make sense of. It is beyond the scope of this paper to examine whether we should let go of the value of individuality. If this were the case, then a different kind of ‘love’ would be a matter of justice – a kind of love that is not personal and therefore is more amenable to control by the lover without loss of significant value. Some readers might be happy with this conclusion.

The wide acceptance of the second claim, that love cannot be a matter of duty, explains why the very idea of having a duty to love is prima facie implausible. Yet, there are good reasons to believe that there is at least a duty to try to create the conditions for everybody to receive sufficient love. While duty does not require that everybody is loved, it does require that we try and create the social conditions conductive to flourishing loving relationships which exclude as few individuals as possible. There may be various ways of promoting inclusive loving relationships. For example it is plausible that we have a duty of justice to create the social conditions that allow as many children as possible to receive adequate, loving care from adults (Brighouse and Swift Reference Brighouse and Swift2014). Lynch, John Baker and others (Reference Lynch, Baker, Lyons, Baker and Lyons2009) argued that personal love is one important metric of justice for children and adults alike, and argued for policies that would facilitate a more egalitarian enjoyment of loving relationships.

Note that recognising a duty to create the social conditions in which loving relationships thrive, including as many people as possible is coherent with the claims (1) that love is a matter of justice and (2) that there cannot be a duty to love. We can imagine, and hope, that, were we to fulfil the duty to realise the conditions under which love could flourish, love would indeed be enjoyed to a more equal extent by different people – and, that, at the very least, nobody’s life would fall below the sufficiency threshold due to a lack of love. But such a lucky outcome would, however, not dissolve the paradox, since the paradox does not stem from an issue of feasibility but from one of conceivability. Unlike a duty to love, a duty to try to love is conceivable. Yet, even full compliance with such a duty cannot ensure that the non-instrumental good of love as defended here is delivered and therefore that full justice is realised. What we ultimately need is to be loved, not merely to live in a society organised to promote inclusive love relationships. And therefore one cannot reduce the claim (1) that enjoying love is a matter of justice to a less demanding claim that (1*) Attempts to create loving relationships is a matter of justice.

Another answer to the question of love and justice that would fail to dissolve the paradox is to say that, in principle, those who do not enjoy their fair share of the non-instrumental good of love are owed compensation. In the case of other goods that ought to be distributed fairly, but which we cannot distribute fairly, we think compensation is the way of realising distributive justice. And indeed, some philosophers do contemplate the possibility of, and recommend, compensation of people unjustly deprived of love. Philippe van Parijs (Reference van Parijs1995) argued that we should compensate individuals unable to find spouses in what I take to be an attempt to recognise the non-instrumental good of long-lasting loving relationships and its import on matters of distributive justice. If love has the kind of importance I assume it has, adequate compensation for leading a loveless life is unavailable (Gheaus Reference Gheaus2009). Whether or not one thinks that compensating people for lack of love is appropriate, the possibility of compensation is not at odds with the claim that love is a matter of justice – to the contrary, it confirms the claim. If enjoyment of love was not a matter of justice in the first place then surely no compensation could be due for the lack of this enjoyment.

The solution I favour is to resist the claim in (3) that all matters of justice are also matters of duty and to embrace the possibility that there can be injustice without perpetrators and without conceivable duties of remedy.