Duty, Virtue, and Filial Love

Abstract

The aim of this paper is to argue that the normative significance of the inner aspects of filial piety – in particular, filial love – is better captured when we understand filial love as part of the virtue of filial piety rather than as an object of duty. After briefly introducing the value of filial love, I argue that the idea of a duty to love one's loving parents faces serious difficulties in making sense of the normative significance of filial love. Then I show why the virtue-ethical approach to filial love, which views filial love as a constitutive part of the virtue of filial piety, can do justice to its normative significance while avoiding the difficulties.

1. Introduction

Many, if not most, contemporary Western philosophers who worked on filial piety have attempted to understand it in terms of duties (or obligations or responsibilities).^{1, 2} That is, their attempts have focused on explicating the nature and grounds of filial duties we have in relation to our parents. On the other hand, some inner attitudes, such as love for one's parents (henceforth, ‘filial love’), are an important part of being a filial child.³ It seems hard to do justice to the normative significance of such inner attitudes if we understand them as an object of duty.

The aim of this paper is to argue that the normative significance of the inner aspects of filial piety – in particular, filial love – is better captured when we understand filial love as a part of the virtue of filial piety rather than as an object of duty.⁴ After briefly introducing the value of filial love, I argue that the idea of a duty to love one's loving parents faces serious difficulties in making sense of the normative significance of filial love. Then I show why the virtue-ethical approach to filial love, which views it as a constitutive part of filial piety understood as a virtue, can do justice to its normative significance while avoiding the difficulties.⁵

2. The Value of Filial Love

There is little doubt that certain outer behaviours – e.g., taking care of one's parents when they are old and infirm, visiting them frequently, and offering financial support if necessary – are important components of filial piety. However, having appropriate inner attitudes (i.e., emotional and motivational states) towards one's parents also seem to be an important component of filial piety. Although people may disagree over the specific contents of appropriate filial inner attitudes, I believe it can be largely agreed that filial piety involves more than outward behaviours.⁶

It seems reasonable to think that being a filial child constitutively involves having appropriate inner attitudes as well as appropriate outer behaviours. If so, what would the candidates be for such filial inner attitudes? As P. J. Ivanhoe claims, plausible candidates for a true basis of filial piety include ‘the sense of gratitude, reverence, and love that children naturally feel when they are nurtured, supported, and cared for by people who do so out of loving concern for the child's well-being’ (Ivanhoe, 2007, p. 299, emphasis added). The intuition that love for one's parents is required for being a filial child is also supported by the proponents of the friendship theory (see, e.g., English, 1992; Dixon, 1995, among others). According to these authors, just as ‘friends are motivated by love rather than by the prospect of repayment’ (English, 1992, p. 149), so would a filial child be motivated by love for helping and supporting her parents. The love desirable for a filial child would be a special kind of love that is distinguished from impartial benevolence. As Keller says, a filial child would ‘do things for the parent willingly, out of love, not out of a motive of duty’ (Keller, 2006, p. 255). It would be reasonable to suppose – and I shall assume this here for the sake of argument – that filial inner attitudes include at least love for one's parents, broadly construed as involving special care and attachment. It is not to deny that the virtue of filial piety may involve other important inner attitudes such as gratitude and respect.

3. Filial Love and Duty

If I am right, filial love is one of the important inner attitudes that are constitutive of filial piety. Being a filial child involves not just performing filial actions motivated by such ‘filial’ emotional attitudes as love, but also having those emotions towards the parents considered independently of their motivational role. Other things being equal, a person who performs filial actions motivated by love for her parents would be more filial – that is, more virtuous as their child in some respect – than one who does the same from a sense of duty, and a person who has filial love would be ‘more filial’ than one who lacks it.

Love for one's parents has special normative significance at least insofar as the parents have loved one and one has accepted and encouraged it in some sense. Of course, if the parents have been consistently unloving – e.g., abusive, cruel, or indifferent – it seems to generate a good excuse for the child not to love them, no matter how great a good filial love is for the parents.⁷ However, if one has received and accepted the parents’ love, one seems to have an additional normative reason to love them. In a similar vein, P. J. Ivanhoe says that the ‘only appropriate response [to the parents’ love] is to keep in mind the nature of their love and, in the warmth of this light, to cultivate reciprocal—yet distinctive—feelings for them’ (Ivanhoe, 2007, pp. 304–5). If so, an adequate theory of filial piety needs to make sense of this normative significance of filial love. That is, a good theory should offer the normative grounds of loving one's (loving) parents to explain why it is morally important for filial piety to be disposed to feel filial love and to be motivated by such love in acting for their benefit.

To do justice to the moral importance of filial love, some philosophers make an appeal to the idea of a duty to love. They claim that a child who has been loved by her parents and who accepted that love has a duty to love them in return. They list this duty among the filial duties. Norvin Richards has offered a view along these lines. According to him, if we have ‘accepted a place in our parents’ affections and encouraged them to continue to love us, when we were young and under their care’, then we have a duty ‘to give our parents a roughly similar place in our own affections’ (Richards, 2010, p. 240). If you have met this condition, he says, then ‘you do owe some degree of affection in return, partly because to have a central place that you welcome in the affections of another person is so great a good’ (Richards, 2010, p. 236). Simon Keller also seems to support the idea of a duty to love when he explains how the duty to be loyal to one's parents involves duties of feeling, including a special concern for them (Keller, 2007, Ch. 6). S. Matthew Liao is another philosopher who argues in favour of the idea of a duty to love in his recent works (Liao, 2006; 2015, Ch. 4).⁸ I will argue that there are serious difficulties in the idea of a duty to love one's parents, as well as in the general idea of a duty to love.

One of the common objections to the idea of a duty to love – or the general idea of a duty to feel a certain kind of emotion – is the commandability objection. According to this objection, loving someone cannot be a duty because having emotions like love towards a particular person is not sufficiently controllable – i.e., not something we can bring about at will – and therefore not commandable. It seems largely agreed that love involves feelings that are not under our control at any given moment and that it may also be beyond our power to cultivate them over time. The point is that what is not sufficiently controllable cannot be a duty at least insofar as the familiar principle of ‘ought’ implies ‘can’ holds.⁹ Barbara Solheim, for example, offers the following version of the commandability objection:

Our ability to love (or to generate love) involves numerous variables, from our temperaments and personalities, to the fund of role models and background experiences we have (or lack), to our ability (or inability) to be fully attuned to others. Although it is true that over time, some who have difficulty loving can learn to love, doing so still involves variables that are to some extent beyond our control. Because of this, love cannot be required as a duty. (Solheim, 1999, p. 14, emphasis added)¹⁰

If she is right, it seems to pose a serious challenge to the idea of a duty to love.

There is an additional difficulty related to controllability that is specific to the idea of a duty to love one's parents. The difficulty is that whether and to what extent a person can love her parents largely depends on how they have treated and educated (or failed to treat or educate) her while raising her. It is to a large degree the parents’ responsibility to raise their child as someone who can love other people genuinely. Richards even says that ‘it is a parent's obligation to enable his or her child to see that love is to be reciprocated’ (Richards, 2010, p. 240). When parents fail to raise their child properly, she may become a person who is sceptical about the value of love or who lacks the emotional resources necessary for loving and caring for someone else. If so, she may grow up to be a person who cannot love another person, including her own parents, through no fault of her own. The main problem here is that the child's being able to love depends on conditions out of her control.

It may be responded that only parents who do not love their child fail to enable their child to love other people, and that this is not a problem for the idea of a duty to love one's parents since a child does not have a duty to love such unloving parents. Richards may respond along this line since he claims that whether you have a duty to love your parents ‘turns […] on how much they loved you: on how central a place you had in their affections […]. [Filial duties] require us to give our parents a roughly similar place in our own affections’ (Richards, 2010, p. 234). However, even if we narrow down the scope of the duty to love to loving parents, the worry remains. For even loving parents may fail to raise their child to be loving. For example, if the parents’ love reaches a level of indulging the child, it may spoil their child so that she grows up as an egotistic person who is incapable of loving and caring about someone other than herself. Therefore, the fact that the parents’ way of upbringing has a major impact on the child's ability to love gives us an additional reason to believe that loving is not sufficiently controllable to be a duty.

Many philosophers, however, hold that we can have control of emotions like love at least to a reasonable degree (e.g., Solomon, 1973; Sommers, 1986; Sherman, 1999; Keller, 2007, Ch. 6; Richards, 2010, Ch. 11; Liao, 2015, Ch. 4). Although they admit that we do not have direct control over our feelings and emotions or the ability to have certain feelings at any given moment at will, they still reject that we are absolutely passive in terms of our emotions and have no control over what we feel. They argue that we can have sufficient control over our emotions like love because our efforts can affect our emotions over time. We can, according to them, at least indirectly affect how we feel through our long-term efforts. For this reason, Nancy Sherman argues that we have some sort of ‘emotional agency’ in this sense: ‘We nurture our capacity for intimate love in the context of ongoing relationships and a will to love in certain ways; we often know how to nip unjustified anger in the bud; we can catch overweening pride and curb it’ (Sherman, 1999, p. 294).

Based on this assumption, Liao argues that the affective aspect of love is sufficiently controllable, and thus commandable (Liao, 2015, pp. 101–13). He suggests various ways in which we can intentionally bring about or promote the emotional aspect of love. According to him, there are at least three ways to bring about particular emotions: giving ourselves reasons to have those emotions, reflecting on the reasons why we tend to have particular emotions in particular circumstances, or deliberately placing ourselves in such circumstances. Suppose that, as Liao argues, parents have a duty to love their children.¹¹ Then, for example, we may bring about our love for our children by giving ourselves a reason to love them, say, that ‘children need this emotional aspect of love in order to develop certain fundamental capacities necessary to pursue the basic activities’; we may reflect on the fact that ‘our antipathy towards the child is due to the fact that the child was unplanned and that the child was born at a time when we already had too many children’; or we may try to ‘have enough sleep each night so that we would be more loving toward the child’ (Liao, 2015, p. 111).

Given the various possible ways to affect our emotional aspect of love, it seems reasonable to acknowledge that our love is controllable at least in some sense. The remaining question to ask is whether love is controllable in the right way to ground the alleged duty to love. Solheim denies that it is: ‘Since our efforts to cultivate an emotion do not guarantee our success, to say that we have a responsibility to try to cultivate an emotion is plausible, but to say that we have a responsibility to succeed is not’ (Solheim, 1999, p. 12). In response, Liao first points out that ‘it is not necessary that in order to have a duty to V, V must be something one can bring about with guaranteed success, supposing that certain reasonable background conditions obtain’ (Liao, 2015, p. 115). For example, he says, while it seems reasonable to say that professional chefs have a duty to make good dishes, no chef can guarantee that their dishes will be good, although they are normally good. Similarly, he argues, a duty to love can exist even if the efforts to bring about emotional aspects of love do not promise guaranteed success.

I agree that guaranteed success in bringing about V is not necessary for there to be a duty to V. However, there is a problem in the attempt to argue in favour of a duty to love based on the possibility of raising the likelihood of loving someone. One serious problem is that a duty to V is being conflated with a duty to cultivate the ability necessary to V. Strictly speaking, the actions aimed at producing an occurrent loving emotion or cultivating the disposition of loving someone are not themselves part of loving. Rather, the fact that one has to put so much effort into trying to love that person may be a sign that one does not love her, at least not yet. Loving is the end that those effortful actions aim at.

Compare belief. Suppose that there is a duty to believe P. Just like the case of love, since we cannot simply decide to believe something at a given moment, this duty cannot be a duty to believe something at a particular moment. On the other hand, we can choose a course of action that will affect our beliefs – we may have a duty to make choices in ways that increase the likelihood of us forming the belief that the duty requires. For example, one can put oneself into a situation in which she is more likely to form the required belief and try to expose herself more openly to the evidence in favour of that belief.¹² If so, if there is such a thing as a duty to believe P, it should mean a duty to do things that tend to raise the likelihood of believing it, rather than the duty to believe it, period. Putting effort into trying to believe P, is not part of believing P. Thus, if there is a duty to believe P unless one actually forms this belief, one entirely fails to fulfil this duty no matter how hard one tried to believe it.

Similarly, as we cannot simply decide to love our parents, what one can do is just to try to love them, that is, doing things that tend to raise the likelihood of loving them. Trying to love includes what a truly loving child would do for her parents such as visiting and contacting them more frequently, reminding herself of good memories they shared, and focusing on positive qualities of one's parents. Or, as Confucians recommend, one can also read stories of children who display filial piety as in the Classic of Filial Piety or engage in rituals such as regular meditation on all the things one's parents have done for one – which, of course, may also cultivate gratitude as well as love for one's parents.

However, such effortful activities do not themselves constitute loving. The important point is that loving is not an activity that consists in putting one's efforts into generating the emotional state and succeeding in it. I suspect that the proponents of a duty to love, including Liao, conflate the alleged duty to love with a duty to produce the disposition to feel loving emotion. Loving begins when the efforts to produce loving emotions end. Once such an emotional disposition is formed, then what the subject does and feels afterwards would begin to constitute her loving for the first time. In other words, it is only after the success of these efforts to love that loving begins. Those efforts are not part of fulfilling the alleged duty to love at all, regardless of whether they succeed in bringing about the targeted emotion. Rather, they are fulfilling the duty to produce the disposition to feel love. I acknowledge that, in reality, most people's love has its ups and downs and thus it is not so mechanical that first you make the effort and then you love. At least sometimes, you love, but you are also making the effort to keep yourself in the position to love in the face of changing circumstances and changes in people. Still, the efforts to keep loving someone, just like the efforts to cultivate the love for the given person, are not themselves part of loving, although they might be important to protect the loving relationship.

4. Filial Love and the Virtue of Filial Piety

I have argued that the idea of a duty to love one's parents faces serious problems. In this section, I would like to argue that a virtue-ethical approach to filial love has a theoretical advantage. I agree with P. J. Ivanhoe that ‘[s]eeing that what is called for is a certain critically informed attitude or state of character shows why filial piety is best thought of as a virtue’ (Ivanhoe, 2007, p. 305, emphasis added). Virtue involves having the right kind of feelings, not just the right kind of actions. A virtuous person acts motivated by appropriate inner states rather than acting mainly from a sense of duty. From a virtue-ethical perspective, feeling appropriate emotions partly constitutes living a good life as a human being, as performing appropriate actions does. As virtue ethics advises us to live a good life, feeling and cultivating the disposition to feel appropriate emotions has normative significance. Thus, I would like to suggest that filial love can be best understood when it is viewed as part of virtue, in particular, the virtue of filial piety.

If we understand filial love as partly constitutive of filial piety as a virtue, then a person who lacks love for her parents for their own sake, or other constitutive emotional or motivational inner attitudes, would lack the full virtue of filial piety. Other things being equal, if one child loves her parents whereas the other doesn't, we can say the former is more filial – or more virtuous as a child – than the latter, at least in that regard. There are two aspects of filial love that need to be explained. On the one hand, although there are some ways to raise the probability of producing filial love, feeling appropriate emotional states seems to be beyond our direct control. On the other, at least insofar as our parents have loved and cared for us, loving them does not seem to just be morally optional. That is, it seems that a person who lacks filial love for one's parents deserves at least some sort of negative moral evaluation or ‘moral criticism’. As Lawrence Blum says, ‘what has moral significance goes far beyond what can be made an object of duty or obligation’ (Blum, 1980, p. 159).

I believe it should be explained why a person who fails to love her loving parents deserves some negative moral evaluation. Let me begin by introducing two different kinds of moral evaluation: the evaluation of blameworthiness (or praiseworthiness) and that of viciousness (or virtuousness).¹³ While one may be blameworthy only for something that is under her voluntary control at least to a reasonable extent, one can be criticized as being vicious or not being virtuous even for what is not under her voluntary control.¹⁴ Based on this distinction, Nicolas Bommarito shows how emotions can be subject to the latter kind of moral evaluation:

Even if we cannot be blameworthy for involuntary states, such states can still make one a morally worse person. Even if someone with involuntary racist emotions is not blameworthy for those emotions, they still reflect poorly on his moral character. He would, after all, be a better person if he did not have those emotions. Learning that someone feels racial contempt, greed, or envy leads us to revise our assessment of their moral character even if we know that they can't help but have such feelings […]. Blame and viciousness are distinct moral evaluations. If nature has made it so that I cannot help but have sexist attitudes, nature has thereby made me vicious. If I truly have no control over my attitudes, I may not be responsible for such attitudes. But that is a separate question; in any case, such attitudes are a blemish on my moral character, and I would be a better person without them. (Bommarito, 2017, p. 90, emphases in original)

According to this view, a certain kind of emotion or lack thereof can be grounds for a negative moral evaluation of the agent's character – i.e., the evaluation of viciousness – independently of whether her voluntary actions or omissions are responsible for having or lacking such emotions. We can apply this point to filial love. If the virtue of filial piety involves loving one's loving parents, then lack of filial love would provide us with the grounds to judge that the subject in question is at least less than fully virtuous – or incompletely virtuous – as their child. (It would sound too harsh to say that a child is vicious for lacking filial love especially if she sufficiently performs filial acts.) Inability to love does not always set the given agent free from moral evaluation, insofar as the basis of the inability is her own character. That is, if the defect in character is the reason for the inability, then the character is subject to moral criticism.

If so, we can explain why it is morally important to have filial love without relying on the idea of a duty to love one's loving parents. Filial love is important because it serves as grounds for the moral evaluation of one's character regardless of whether having it is within one's control. Julia Driver also lends support to this point when she says that ‘a child may have not duty to love a parent (as opposed to a duty to feel gratitude), but failure to love under some circumstances can reveal something very bad about a person's character that may still warrant negative evaluation’ (Driver, 2014, p. 9) (although she does not further elaborate on this point). It is true that emotional states unexpressed in actions are hard to detect. I think this is why it is relatively hard to observe a case where someone morally criticizes another for merely having such emotions.

However, moral criticism of having or lacking a certain kind of emotion is part of our common experience since we often morally criticize ourselves. For example, if we have an emotion that we judge to reflect viciousness, we tend to feel shame or even self-disgust. The issue is not necessarily about whether these emotions are likely to lead to any wrong actions. The criticism is made just in respect of feeling an inappropriate emotion or lacking an appropriate one. For example, one may feel ashamed of not feeling grateful to one who has been supporting her both financially and materially, while acting as a grateful person would act, though out of a sense of duty. Similarly, one may feel ashamed of not feeling loving emotions towards one's parents.

When we are ashamed of ourselves for feeling or failing to feel a certain kind of emotion, the object of shame is not necessarily what I have done (or failed to do) to bring about or eliminate that emotion. It is about the fact that the emotion is felt out of one's character, which speaks badly of one's character. Whether a person is subject to moral criticism for X is not entirely determined by what is within her control or whether X is voluntary. It only distinguishes whether one is subject to the evaluation of blameworthiness or that of viciousness, both of which are moral evaluations. Rather, whether a given person is subject to moral criticism for X is determined by whether X reflects her value commitment or not. If a person lacks what a virtuous person would have due to some factor that has nothing to do with her value commitment, then she would not be subject to moral criticism. For example, if a person fails to feel grateful to her benefactor just because she is an ungrateful person, then she is still to be criticized for being vicious, in particular, for being ungrateful. In contrast, if the agent fails to feel grateful because of her mental illness related to memory loss, then she would not be subject to comparable moral criticism.

It might be argued that filial love's role as grounds for a negative moral evaluation, in particular, that of viciousness, does not adequately capture its normative significance. This sort of evaluation, it might be claimed, is not very different from merely aesthetic evaluation in its normative force, given that it loses its connection with the agent's voluntary agency, which seems to be central to anything that has normative significance. The claim may be that, although both evaluations of blameworthiness and that of viciousness are both value judgements, the latter, like purely aesthetic evaluation, does not have sufficient normative implication. For example, the judgement that someone is not beautiful is a value judgement but does not necessarily imply that the agent ought to do something to make some change about it. Similarly, it might be argued, the judgement that due to a lack of filial love someone is less than fully virtuous as someone's child does not necessarily imply that she ought to do something about it, especially in the cases where the agent is incapable of loving her parents.

My first response is that having or lacking a certain kind of emotion is relevant to our evaluation of the agent's moral character, while her appearance is not. That is, although neither kind of value judgement presupposes the agent's voluntary control, the evaluation of viciousness seems to fall under the category of moral character while the evaluation of beauty does not. I believe that the view that morality only concerns aspects that are under our voluntary control is a myth, given that our life and character involve much more than what we do through voluntary agency.

Moreover, if filial love is understood as a constitutive part of virtue, in particular, of filial piety, then there is a sense in which we can say that filial piety demands the agent to love her parents. Thus, if the agent fails to respond to this demand of filial piety by failing to love her parents, she is subject to a kind of negative moral evaluation. I believe this is how the virtue-ethical approach can do justice to the normative significance of filial love without relying on the idea of a duty to love.

Let me clarify what I mean by saying that a virtue demands something. By saying that a virtue V demands an agent A to X in a given situation S, I mean that an agent with the full virtue of V would (characteristically) respond to the situation S by X-ing. This understanding of a virtue's demand is inspired by Julia Annas, who understands the guidance in terms of virtue as ‘guidance as demand. That a certain action is the brave, or considerate or generous thing to do creates a demand that it be done, and if it is the action that I am in a position to do, it creates a demand that I do it’ (Annas, 2015, p. 611). Annas illustrates how a virtue makes a demand as follows:

Suppose, to give an example, that I come across picnic litter in a scenic place. I do not want to pick up the litter; I take it that nobody does […]. [T]o the extent that I am a considerate person I will feel that picking up the litter is the right thing to do, that I ought to do it, should do it. (Annas, 2015, p. 611)

In this example, according to Annas, ‘[the virtue of] considerateness demands that I pick up the litter, which is why it is the right thing for me to do’ (Annas, 2015, p. 614). This is what she calls a ‘demand of virtue’, and I think it can be extended to cover what the agent feels, as well as what she does. Filial piety demands that I feel love and gratitude to my loving and caring parents. If I fail to feel these emotions towards my parents, then I would be less than ‘fully filial’ in that regard, just like when I fail to act to help and support them.

Note that the demand of virtue does not necessarily imply that it is felt as a demand by the agent. A fully virtuous person feels it almost as if it were her ‘second nature’, and, as Annas points out, it will ‘normally be felt as a demand by the learner and the incompletely virtuous’ (Annas, 2015, p. 612). If an agent fully possesses the virtue V, then she would not feel the demand of V as a demand. Then a person who possesses the virtue of filial piety fully, who loves her parents already, normally wouldn't feel loving them as demand and wouldn't even think about the possibility of not loving them.

One advantage of the virtue-ethical approach to filial love is that we can make sense of its normative significance appropriately considering how virtuous the given agent is. That is, virtue ethics makes a different command according to the level of the virtuousness of the agent in question. Virtue ethics commands the agent to do whatever she can to cultivate and exercise her virtue. That is, what it commands the agent is to do what one can do at the moment to get as close as possible to living virtuously. Thus, both a non-loving child who struggles to love her parents and acts as if she does and a loving child who supports her parents out of love are doing their best to achieve the ideal of living as a child who possesses the virtue of filial piety. Given the difference in the ability to love their parents at that moment, it is inappropriate to blame the non-loving child for failing to fulfil any duty or for doing anything wrong, insofar as she is doing her best under the circumstances. This is why I think the virtue-ethical approach to filial love can avoid the commandability objection.

Virtue demands a non-virtuous person to act as a virtuous person would act. There are two important points in this demand. First, a non-virtuous person can become more virtuous by imitating a virtuous person's act, since, as Aristotle famously says, a person becomes virtuous by performing virtuous actions: ‘we become just by doing just actions, temperate by doing temperate actions, brave by doing brave actions’ (Aristotle, Nicomachean Ethics, 1103b1–2, 1999). In this sense, acting as a loving child would act is an effective way for a ‘not-yet-fully-filial-person’ to cultivate filial love. Acting as a loving child would act, unlike actually feeling the emotion of filial love, is something even a non-loving child can do.

Moreover, acting as a loving child would act can count as meeting the minimal demand to perform actions beneficial to our parents, although those actions would have had more ethical value if they were motivated by filial love. Benefiting or helping one's parents is an important behavioural aspect of filial piety, which has a significance that goes beyond the mere imitation or cultivation of virtue. For example, actions such as offering financial support or medical care to our parents have moral significance independently of the underlying motivational state. Thus, for a person who is not able to feel love or be motivated by love in benefiting and helping her parents at the time of action, performing those actions from any available motivation for her at that time would be the best she can do to get closer to living as a fully virtuous person would.

Thus, from this perspective, virtue ethics can advise even a non-loving child to try to love her parents and act as a loving child would act. This is true regardless of whether she can actually love them since she can at least try to love them or act as if she does. In the case of a fully virtuous person, virtue ethics tells her to exercise the virtue she already has. It is likely that this demand of virtue to feel and act appropriately would not even be perceived as a demand since her inclination is in line with what it demands. If a person already has the full virtue of filial piety and thus loves her parents, then there is no need for any further normative requirement, since all she ought to do is to feel and act from her character.

Even if there is such a thing as a duty to love, it seems to sit very awkwardly with both non-loving and loving children. For a non-loving child, it would be to require something she cannot do, at least at a given moment, thereby rendering the requirement of duty futile. For a loving child, on the other hand, it would be requiring something she is already doing or inclined to do (or feel), thereby rendering the requirement of duty redundant. It is interesting that neither a non-loving nor a loving child can simply decide to love or not to love at will, though in different senses. A non-loving child cannot simply summon up loving emotions at will. A loving child cannot simply remove this emotion at will. In this sense, love is non-voluntary even if producing the disposition to love may be voluntary to some extent.

To do justice to the value of filial love, we need to shift our focus from rightness of actions to virtuousness of character. If we approach filial love in terms of the idea of a duty to love, we would end up focusing on what we can do to bring about filial love. I think the discussion of what a child can do to love her parents is a red herring. As mentioned above, whether or not a child loves her parents heavily depends on how her parents have raised them. So, the discussion of filial love goes from what the child can do to bring it about, to what the parents can do to bring it about and deserve it. If the parents are abusive and indifferent, for example, it would be hard for her to love them even if she tries hard. That is, the way parents raise their child can deeply affect the scope of filial piety that can be controlled by the child herself. Therefore, to cover the ethics of the wide range of the relevant parties, the discussion should focus on the virtuousness of filial love, and then spread out to ask what all the relevant agents – which include not just the child herself but also the parents, educational institutions, and the government, etc. – can do about it.

My suggestion has been that lack of filial love deserves negative moral evaluation not because it signifies a failure to fulfil one's duty, but because it reflects on one's defective character. In my view, a person is less than fully virtuous as a child to the extent that she lacks filial love for her parents. This point can be made independently of whether she is capable of loving her parents or not; regardless, it is an important part of being a filial child. I have said that it is inappropriate to blame a child for not loving her loving parents insofar as it is through no fault of her own. She has done her best to be a filial child towards her parents but failed because it was beyond her control, and thus she is not responsible for the lack of filial love. As she does not have the full virtue of filial piety, what she can and ought to do at this moment is to try to love rather than to love her parents.

Note that I am not claiming that there are no such things as filial ‘duties’ at all. I admit that there can be various kinds of duties and rights involved in a parent-child relationship. My view is just that, even if filial duties or parental rights exist, they at best serve as ‘fallbacks’ or ‘constant constraints’ as understood in Jeremy Waldron's discussion of the marital relationship (Waldron, 1988, p. 629).¹⁵ Waldron admits that love and affection are often sufficient enough to generate desired acts between married partners. However, he says, rights, and thus the correlative duties, as spouses are still necessary as ‘fallbacks’ in case the relationship collapses. For example, suppose that a husband does not love his wife, who has not enough income to pay for her basic living expenses. If he did not have a duty to offer minimal financial support to his wife and if she did not have a right to the support, then she would be vulnerable to poverty caused by her husband's whim. In this sense, the relevant duties and rights serve to constrain the actions of the spouses to prevent the worst-case scenario in which they not only lack love and affection but also omit actions minimally required of legal partners.

A similar point can be applied to a parent-child relationship. Filial duties and parental rights are no more than fallbacks in the sense that they come into play only when the parents or the child or both lack the desirable emotional and motivational attitudes – e.g., love and concern – and thus their intimate relationship collapses. I suspect that the majority of contemporary Western theories of filial piety are focused on filial duties because the main motivation behind them is to discover what adult children owe to their old and infirm parents as the moral basis for legislation and public policy related to how the adult children ought to treat their parents. Such a duty-centred approach fails to capture the emotional and affective aspects of filial piety such as love and gratitude, which are a crucial part of an ideal parent-child relationship. It is this kind of inner attitudes that distinguishes a parent-child relationship from a mere creditor-debtor relationship or benefactor-beneficiary relationship, and that gives the former a distinctive kind of value.

Also, I do not deny that there can be duties somehow related to love. While I denied a duty to love, I have admitted that there may be duties such as a duty to try to love and a duty to act as a loving child would act. There being such duties is not incompatible with my view on filial love. Unlike the duty to produce filial love, the duty to try to love can be fulfilled even if the agent ends up failing to love, insofar as she does her best to love the object in question. This duty can be understood as part of the duty to pursue the moral ideal of filial piety as a virtue. In general, part of being virtuous is to have appropriate emotional and motivational states, in addition to acting in an appropriate way. These duties would not be subject to the commandability objection, since even a person who does not or even cannot love her parents may still try to love them or act as if she does. This point is also compatible with the intuition that we could not be blamed for failing to love certain people after having made the utmost effort to love them. Those duties naturally invite us to view filial love as part of virtue. A virtue-ethical approach normally suggests what an ideally virtuous person would be like and sets becoming like them as a goal to pursue. If an ‘ideally filial child’ loves and cares for her parents and acts accordingly, then what the agent ought to do, according to the virtue-ethical approach, is to do whatever it takes to get closer to live like them.

5. Conclusion

In this paper, I have argued that filial love is morally significant and that its nature and value are better understood as a part of the virtue of filial piety rather than as an object of duty. Filial love, along with other inner attitudes such as gratitude and respect, would be an example of what is morally valuable but cannot be a proper object of duty. I believe that the normative significance of such inner attitudes can be best understood when they are considered as a constituent of virtue. If I am right, this lends support to the view that important constituents of filial piety are better captured by understanding filial piety as a virtue rather than merely as a source of duties to one's parents. It is morally important to love our parents, but it cannot be a duty.¹⁶