Caring intuitively involves a commitment on the part of a caregiver to the subject of her care. Indeed, as I shall argue in this article, it is hard to see how caring could fail to involve some such commitment. Given the intuitiveness of this claim, it is striking that care ethicists often fail to discuss caring relations explicitly in terms of the sort of commitment that they engender (this is true, for example, of Reference BowdenBowden 1997; Reference HeldHeld 2006; Reference EngsterEngster 2007; Reference PhillipsPhillips 2007; Reference SloteSlote 2007). Even when commitment is explicitly identified as a part of what it means to care, this is not typically accompanied by any sustained critical discussion of the ethical significance of that commitment (for example, Reference NoddingsNoddings 1984, 24; Reference Friedman, Hanen and NielsonFriedman 1987, 105–107; Reference Tronto, Jaggar and BordoTronto 1989, 173; Reference Meyers and CalhounMeyers 2004, 292). Perhaps this should not be surprising, given that the concept of commitment per se has received relatively little in the way of philosophical analysis (Reference ShpallShpall 2014, 146). In any event, a detailed examination of caring considered explicitly as a form of commitment has the potential to shed interesting new light on our understanding of care ethics.

In this essay I explore just one of the implications of a commitment‐centric conception of caring, arguing that such a conception ultimately calls for us to endorse moral partialism. Many care ethicists have already expressed support for moral partialism, but my approach goes beyond arguing that partialism ought to be endorsed by care ethicists. Rather, my claim is that moral partialism is a necessary implication of any care‐ethical view, in virtue of the central importance of commitment to caring.

I begin in section I by making some brief remarks on the nature of caring, to which I will return in later sections of the article. In section II I discuss the nature of commitment, and show why caring ought to be understood as a species of commitment. In section III I discuss the philosophical attractions of moral partialism, and critically consider two influential arguments that have been used to defend it. I then show, in section IV, that my commitment‐centric account of caring vindicates moral partialism. I also show that it avoids the difficulties associated with the arguments for partialism discussed in section III. I then reply to some objections to my account in section V, and briefly consider some of the wider implications of my argument in section VI.

I. On Caring

Caring is an activity aimed at meeting the needs, or promoting the well‐being, of a cared‐for. Caring is thus a goal‐oriented activity, and individual actions undertaken as a part of caring will aim to bring about certain specifiable results: the prevention or cessation of unnecessary suffering; recovery from a particular illness; or healthy growth and development. Archetypal instances of caring take place between a responsible caregiver and a dependent cared‐for, such as a nurse and her patient or a mother and her child. As these examples show, it is possible broadly to distinguish between two different sorts of caring relationship: those that involve someone acting as caregiver in a professional capacity (as in the nurse–patient relationship), and those that do not (as in the mother–child relationship). Following the literature, I will refer to these types of relationships as instances of formal and informal caring, respectively. The distinction between formal and informal caring will be discussed in more detail in the next section, specifically in relation to the types of commitment that they embody.

Of course, caring may be a characteristic of many sorts of relationships, such as those between friends, siblings, and lovers, and the archetypal examples of caring just mentioned should not lead us to overlook this. It should also be noted that individuals may at the same time occupy both the role of caregiver and that of cared‐for (as shown by the discussion of “nested dependencies” in, for example, Reference KittayKittay 1999 and Reference Ward, Barnes, Brannelly, Ward and WardWard 2015). Finally, it makes sense to say that caring may also be directed toward the environment, or to abstract concepts such as economic justice, or the alleviation of suffering.

Caring in these latter cases seems more accurately described as caring about rather than as caring for (following Reference NoddingsNoddings 1984, 21). Nevertheless, a fully spelled‐out conception of caring ought to be able to accommodate instances both of caring about and of caring for. This is one of the strengths of the definition of caring put forward by Berenice Fisher and Joan Tronto. Fisher and Tronto argue that caring is “a species activity that includes everything that we do to maintain, continue, and repair our ‘world’ so that we can live in it as long as possible” (Reference Fisher, Tronto, Abel and NelsonFisher and Tronto 1990, 40). This definition is often rejected as too expansive, requiring us to identify as instances of caring many activities that do not seem to fall within its domain (for example, Reference HeldHeld 2006, 31; Reference EngsterEngster 2007, 24). Still, without endorsing it, we may usefully hold up Fisher and Tronto's definition as one of the ways in which different types of caring can be unified under a single conceptual schema.

While it is necessarily a form of activity, at the same time caring also has a psychological or affective component. Accordingly, for an activity to constitute an instance of caring it must not only be aimed at meeting the needs or promoting the well‐being of a cared‐for; it must also be motivated by a benevolent concern for the well‐being of the cared‐for (for example, Reference NoddingsNoddings 1984, 9; Reference HeldHeld 2006, 32–33; Reference SloteSlote 2007, 10–12). Someone whose actions meet certain caring requirements cannot genuinely be said to care if her actions are not, at least in part, the product of a concern for the cared‐for. If such an agent acts as she does only because she feels obliged to do so, or because she anticipates being rewarded in some way for her efforts, then she cannot be said to be exercising caring agency.

Finally, it should be noted that caring is an essentially normative concept. Thus, to say that an agent is engaged in caring presupposes that her actions are at least minimally constitutive of care. Any would‐be caregiver whose actions persistently fail to meet the needs, or to promote the well‐being, of her cared‐for, or who actively harms her cared‐for, cannot be described as providing care. Of course, it is possible to speak of good and bad, and better and worse, care. As I see it, then, care is a threshold concept: once a minimally satisfactory standard of care has been provided, it is possible to speak of caring as taking place. Any form of activity that fails to meet that standard, however, does not qualify as care, despite the morally praiseworthy motives that may underlie it. Of course, what any caregiver will aim at providing is the best possible care, and not just minimally satisfactory care. What constitutes the best possible care, and where the threshold at which genuine caring begins, are difficult questions that cannot be answered in the abstract. Rather, the answers to them must be given on a case‐by‐case basis, and will depend on a number of factors. These include the resources that the caregiver has at her disposal; the context in which care is being provided (for example, whether at home or in an institution; whether by a friend or by a spouse, or some other relative); and the cared‐for's experience of that care (for example, whether as attentively and compassionately provided, or as resentful and reluctant). Such considerations point to the complexity of caring relationships, and to the need to evaluate each such relationship on its own terms.

II. Caring as Commitment

As will be seen in this section, caring requires a commitment on the part of a caregiver. Before showing why this is the case, however, it will be useful to begin by making a few general remarks on commitment itself.

In part, to describe someone as committed is to say that she reliably or consistently acts in ways that promote goals or ends associated with her commitment. Someone committed to the election of a particular political party will thus consistently vote, campaign, or fundraise for that party. Commitments are about more than mere outward behavior, however. This is because such behavior might be the product of coercion, habit, or preference, rather than commitment. (Having a preference is not the same as being committed. To see this, consider my having a preference for a particular brand of coffee: no matter how enduring this preference may be, it does not make me committed to the brand in question. This is because I would stop buying that brand tomorrow if I found an alternative that I enjoyed even more.)

Commitments may be distinguished from preferences, habits, and coerced behavior by considering the psychological attitude underlying them. Commitments, that is, are grounded in an agent's values. They are formed in response to the agent's judgment that the object of her commitment is good, important, or otherwise worth pursuing, perhaps (though not necessarily) for its own sake. Thus a conservationist is committed to protecting an endangered species because she takes its continued existence to be good, either in its own right, or as a means to the preservation of the wider ecosystem. Similarly, a mother is committed to her child because of her love for it, and because of the importance that she sees in its continued well‐being.

Now, caring may also be described as a form of consistent behavior, and as being engaged in as a response to the importance of the cared‐for and her well‐being. Caring behavior must be consistent because any would‐be caregiver who only occasionally attempted to meet the needs or to promote the well‐being of a cared‐for could not be described as genuinely caring. This follows from the account of caring provided in the previous section. An agent's lack of consistency in attempting to meet the needs of a cared‐for signifies either a lack of concern for the cared‐for, or the agent's unwillingness to act on that concern. Inconsistency in either of these respects is enough to show that genuine caring is not taking place. Of course, some accounts of caring differ from the one given in the previous section, in that they see caring as constituted by either an appropriate motive (for example, Reference SloteSlote 2001; Reference Slote2007) or by the meeting of the cared‐for's needs (for example, Reference BubeckBubeck 1995), and not, as I have claimed, by the conjunction of both of these. Even on such accounts, however, consistency is still necessary if we are to identify the relationship as a caring one. An agent who only occasionally acts so as to meet the needs of another, or who only occasionally feels a concern for another's welfare, can hardly be said to care for that other. Thus, commitment is a necessary aspect of caring.

At this point it might be thought that my discussion of commitment has counterintuitive implications concerning the sort of actions that we are entitled to describe as instances of caring. The actions that I have in mind here are those that might be described as isolated acts of beneficence: pulling a stranger out of the path of an oncoming bus; buying a meal for a homeless person; helping a child lost in the supermarket to find its parents or guardians. None of these actions seem to involve a commitment to those benefited by them, yet surely we wish to refer to them as instances of caring. In fact, my account does allow us to describe isolated acts of beneficence as instances of caring, as I will now show.

It is plausible to say that in each of the examples given above, there is no commitment on the part of the agent to the persons that her actions benefit. Commitments are more durable than single, isolated instances of beneficence. If the agent in question were really committed to this particular lost child, then she would wish to continue to care for it in some way even after reuniting it with its parents or guardians. She would have an ongoing concern for the needs and welfare of that child. Yet of course this is not the case, and any such concern would almost certainly be rejected as inappropriate by the child's caregivers. Once our agent has seen the child safely back with its caregivers, she rightly sees her relationship with it as being at an end. Similarly, buying a homeless stranger a meal does not demonstrate an agent's commitment to that particular homeless person. If an agent were committed to the well‐being of a particular homeless person, we would naturally expect that commitment to extend beyond the buying of a single meal. It would likely also include, for example, ensuring that the person in question has a safe place to sleep for as long as it is needed. But the beneficent stranger in our example has no intention of embarking on any such lengthy course of action, despite knowing the dangers associated with sleeping rough.

None of this means, however, that beneficent actions of the sort described above are not the product of any form of caring commitment. As noted in section I, we do not care only about particular others: we also care about more abstract goals and values, such as environmental conservation, economic justice, and the alleviation of suffering. Thus, an agent who acts because of a benevolent concern to alleviate suffering may still be said to be engaged in caring, so long as the results of her action cross the threshold that sets the minimally satisfactory standard of care in that context (for example, by actually alleviating some degree of suffering, instead of inadvertently increasing the amount of suffering in the world). Accordingly, if the isolated acts of beneficence described above are the product of some such benevolent concern, as seems plausible, then they may still be described as genuine instances of caring. Of course, in keeping with the analysis of caring given above, if an agent is to be said to care about the alleviation of suffering, she must be sufficiently committed to that goal. Furthermore, in the absence of any such caring commitment, isolated acts of beneficence may not be described as genuine instances of caring. In such cases, it seems more accurate to explain such actions not as the products of genuine caring, but as the products of contingent psychological phenomena, such as a desire to avoid the feelings of guilt that would be prompted by doing nothing to help a person in need.

What these remarks show is that identifying the object of a caring commitment is not always straightforward. Take, for example, instances of formal caring. It is intuitive to think that formal caring involves a commitment to a particular cared‐for, in the same way as informal caring does. However, there seem to be important differences between the types of commitment to which these relationships give rise. This suggests that formal caring primarily (though not exclusively) involves a commitment to the promotion of well‐being generally, rather than to the individual welfare of particular cared‐fors (much in the way that the beneficent agent in the previous example was seen as being committed to the alleviation of suffering, rather than to the well‐being of a particular homeless person).

To see why, notice that nurses may end their shift, go on holiday, or change career without thereby violating a commitment to their cared‐fors (assuming that those cared‐fors are not thereby abandoned). It would be unjustified, that is, for their cared‐fors to rebuke them for doing so. Informal caring is different insofar as it requires a more lasting, and nondelegable, commitment between the caregiver and her cared‐for (with the important exception of certain forms of respite care that the caregiver may be in need of). This is not to deny that many instances of formal caring typically do (and should) involve some degree of commitment to individual cared‐fors. It is to point out, however, that such commitment is generally less deep than those found in informal contexts, and that this is typically not seen as a shortcoming of formal caring.

Of course, formal caregiving may be entered into because of any number of commitments. For instance, some nurses will have a commitment to care for their family, which professional nursing enables them to live up to (at least in part) by providing them with a stable income. Nurses may also have a commitment to their profession itself, insofar as they see nursing as a valuable and important public service, and are dedicated to ensuring not only its continuation, but also the welfare of those working in it. The strength of these commitments will of course vary from person to person, but they each may play an important role in guiding the activities of formal caring. It is thus unnecessary to identify a single motivating commitment held by all formal caregivers per se.

III. The Attractions of Partialism

Before showing how my commitment‐centric account of caring leads to the endorsement of moral partialism, it will be useful briefly to say something about the partialism debate itself. I cannot hope to provide a detailed discussion of all of the relevant arguments here, and will therefore limit myself to explaining partialism's appeal, and to discussing briefly two influential arguments that have been made in its defense.

Moral partialism is the view that we have stronger moral obligations, or more demanding moral responsibilities, toward those near and dear to us than we do toward those distant from or unknown to us. This claim has been found controversial, with impartialists pressing the point that everyone is entitled to the same basic degree of moral respect. Peter Singer, for example, has argued that to claim that we have less of a duty to people simply because they are personally unknown to us is to be influenced by a consideration that “makes no moral difference” (Reference SingerSinger 1972, 231). Thus our moral thinking ought, on Singer's view, to be strictly impartial.

Notwithstanding the sorts of concerns raised by its critics, moral partialism has seemed attractive to many. This is because it allows us morally to endorse acting upon bonds of love and affection. For example, it seems morally permissible for me to prefer to save my partner or child from a burning building simply because of my relationship to them, and irrespective of the fact that had I left them to fend for themselves I might have been able to save a greater number of people personally unknown to me. Less dramatically, it seems morally permissible (indeed, praiseworthy) for me to spend time and money taking my son out on enjoyable day trips, or buying my partner a birthday present. This is despite the fact that, in both these examples, my time and money could instead have been used to confer greater benefits upon distant others; for example, by donating it to a charity providing famine relief or clean drinking water to those in the developing world.

Some efforts have been made by philosophers working outside of the ethics of care to accommodate partialism within a wider, impartialist framework (for example, Reference BaronBaron 1991; Reference BarryBarry 1995). Such attempts appeal to a distinction between first‐ and second‐order levels of moral justification. Here, first‐order justifications may invoke partialist considerations, but only because of second‐order, impartial rules that license that partialism. For example, such an approach permits agents to have, and to act upon, a greater concern for the welfare of their own children. It does so, however, only because of the existence of a second‐order rule that may be expressed impartially, such as “it is permissible for a child's primary caregivers to have, and to act upon, a greater level of concern for their child's welfare than they do for the children of others.” (This rule is impartial because it applies in equal measure to all children and their caregivers, unlike the first‐order principle that is formulated explicitly with reference to the agent's own children.)

Such an approach generates a number of difficulties, however. First, as argued by Virginia Held, it is likely that first‐order principles licensing partialism will often come into conflict with second‐order ones prohibiting it (Reference HeldHeld 2006, 79). When this occurs, it seems inevitable that an impartialist resolution to the conflict will be the one that a two‐level theory must recommend. For example, if it becomes apparent that a set of first‐order rules are “inconsistent with the [impartial] categorical imperative, it is unlikely that, to the Kantian, we should ever follow them” (79). If two‐level approaches are thus forced, by their internal logic, consistently to prioritize their impartialist content over the partialism that they seek to accommodate, then they cannot deliver the results at which they aim.

A second problem with two‐level approaches is that they give rise to what Michael Stocker describes as a problematic “split between one's motives and one's reasons” (Reference StockerStocker 1976, 454). That is to say, on two‐level approaches it seems morally inappropriate for an agent to be motivated by considering the norm that makes her action morally permissible (or good). To see why this is the case, consider an agent who decides to take her child out on an enjoyable day trip. It is appropriate for this decision to be motivated by her love and affection for her child. Yet this affection is not what makes her action a good one. Its goodness is instead conferred by the relevant second‐order rule, which impartially prescribes that parents or caregivers spend some of their free time engaging in enjoyable activities with their children. An agent who was motivated to spend enjoyable time with her children only because she had spent time considering this second‐order rule, however, would rightly be thought to have acted on the wrong reason; she really ought to have been motivated directly by her affection for her child. A two‐level approach to moral partialism thus alienates agents from the moral reasons for their actions. On balance then, two‐level approaches do not seem to be a promising way of accommodating moral partialism.

The ethics of care is sometimes thought to lead directly to partialism in virtue of its emphasis on contextualism and the need to understand and respond to the particularity of individual cared‐fors, along with its skepticism of universal, generalizable principles (this seems to be the position taken by Reference HalwaniHalwani 2003, for example). Yet the rejection of generalizable rules in favor of a contextual or particularist approach to moral thinking does not entail the truth of partialism. If particularism is true, then what constitutes good care in one context may not be constitutive of good care in another context, and there are no universally valid rules to which we can appeal when deliberating about how to provide care. However, it does not follow from this that we have a greater duty to care for some than for others, which is the claim that partialism endorses. An alternative foundation for partialism must therefore be sought.

Some care ethicists have taken a different approach, arguing for moral partialism by claiming that we are less able to care for distant others due to a lack of adequate knowledge concerning how to provide that care. Given that misdirected efforts to care may actually end up harming those whom they are intended to benefit, it is better for us to focus on caring for those near and dear to us, whom we may be sure of benefiting by our actions. This line of thought has been endorsed by, for example, Daniel Engster and Nel Noddings (Reference NoddingsNoddings 1984, 112; Reference Noddings2002, 22–23; Reference EngsterEngster 2007, 182). Thus Engster writes that when we are faced with epistemic limitations that prevent us from seeing how to provide good care for distant others, or when providing such care risks encouraging dependency, “we might be best advised to do nothing at all” (Reference EngsterEngster 2007, 182).

This is a plausible line of argument, which I do not wish to reject. Yet such an approach does little, it seems, to accommodate instances of moral partialism of the sort described at the beginning of this section. In the burning‐building scenario, for example, I surely know how to provide good care for the strangers at risk, and that is by saving them from the encroaching fire. Similarly, although I personally may not know how best to provide care for those affected by drought or famine, I may reasonably believe that an international aid organization is better epistemically situated than I am. Those working for it do know how to provide good care, and I know this to be the case because they have a publicly accessible track record of doing precisely that. Thus I can be confident that donating my time or money to that organization will help to promote good care, and will also be productive of a greater amount of good than spending the same amount of time and money on buying a birthday present for my partner. Accordingly, if we nevertheless wish to accept that in these scenarios it is morally permissible for me to save my partner and child from the burning building, or for me to buy my partner a birthday present, we need an argument for partialism that goes beyond the one endorsed by Engster and Noddings. I provide just such an argument in the next section.

IV. From Commitment to Partialism

As will be recalled from section II, a commitment may be understood as the consistent promotion of a goal or end, prompted by the value that an agent attaches to that end. Caring, it was there argued, is a form of such commitment. An agent cares for another when she values the well‐being of a cared‐for, and for this reason consistently feels, and acts appropriately on, a benevolent concern for that well‐being.

Commitments, of course, do not arise ex nihilo. Rather, they are the products of meaningful, formative events in our personal history. For this reason, our commitments are often partly constitutive of our self‐conception, whether they are commitments to individuals or to valued projects. Thus we conceive of ourselves as being (among other things) parents or siblings, authors, or human‐rights campaigners. Commitments to persons may be the product of a shared history of mutual support and trust (as with close friends and partners), or of a history of dependency and nurturance (as with infants and their caregivers). Commitments to nonpersons, on the other hand, may be the product of some deeply important personal experience. Thus, a commitment to raising awareness of a rare genetic disorder may result from learning that a family member has been diagnosed with that disorder. Of course, commitments may be formed in any number of ways besides the ones just mentioned. The point, however, is that our commitments are typically a reflection of who we as individuals are; they are the product of our lived experiences and of our interactions with those around us.

Now, when we say that we are committed to a cause or a person, a part of what we mean to convey is that we have a proportionally greater attachment to or concern for that cause or person. Indeed, this greater degree of proportional concern is a part of the concept of commitment. The objects of our commitments necessarily engage our time and efforts more fully than do other forms of activity to which we are not committed. If we say that we are committed to someone or something, yet do nothing in the service of that commitment beyond what we would ordinarily do in the service of things to which we are not committed, then we have spoken falsely. This conclusion cannot be avoided by saying that we may be equally committed to everyone or everything. This is because to talk in this way is equivalent to saying that we are not truly committed to anything. Noddings expresses a thought similar to this one when she writes: “If by ‘I love everyone’ I mean that I would not without just cause harm anyone, that is acceptable …. [But if] that is all I mean when I say that I love my child, or my husband, or my student, then each of these has, I think, been cheated” (Reference NoddingsNoddings 1984, 112). Noddings is referring to love here, but her remarks are no less true when applied to the concept of commitment. If we are truly committed to specific loved ones, or to specific goals, ideals, or projects, then that commitment necessitates that action taken on their behalf be prioritized above actions taken toward persons, ideals, and so on that are not the objects of a corresponding commitment. If there is no such hierarchical prioritizing, then our purported commitments are not genuine ones.

Now, given that caring is to be understood as a specific type of commitment, it follows that caring requires us to prioritize the good of those for whom we care above the good of others for whom we do not care. If there is no such hierarchical prioritizing, then we cannot genuinely be said to be engaged in the practice of caring. Caring, then, requires us to do more for some than we do for others. More formally, the commitment‐centric analysis of caring that I have developed here entails that we have a greater moral responsibility toward those for whom we care than we do toward those for whom we do not care, and to whom we are not committed. This claim, of course, is simply the thesis of moral partialism discussed in the previous section. When caring is considered explicitly as a form of commitment, then, it can be seen that caring is necessarily partial. Accordingly, a conception of morality centered on caring must endorse some form of moral partialism.

It is worth noting here that this argument avoids the objections discussed in the previous section. First, it clearly is not a form of two‐level theory, so conflicts between first‐ and second‐order norms cannot arise. Second, the partialism that I have argued for is a product of our caring commitments. As such, there is no tension between the normative reason that permits (or requires) a caregiver to act in a morally partial way toward her child, and the psychological motive from which she acts. Thus a caregiver promotes her own child's well‐being over that of a stranger because she is more strongly motivated to care for her own child, and the nature of that caring relationship (qua commitment) is what makes it morally permissible for her to act in this way. Accordingly, there is no disconnection between an agent's motives and her reasons for action. Third, my approach licenses moral partialism without relying upon epistemic limitations on our ability to provide care. Thus, even in situations in which I do know how to provide care for those unknown to me, it may be permissible for me to prioritize my already existing caring commitments over the provision of care to strangers.

V. Objections and Replies

In this section I discuss a number of objections that may have occurred to the reader. In the first two subsections below, I consider some general objections to the ethics of care, both of which are already to be found in the literature. My purpose in so doing is simply to show that my account is no more vulnerable to these objections than are discussions of caring that do not emphasize commitment. In the third subsection I discuss an objection aimed specifically at the account of caring for which I have argued.

VI. Wider Implications

Understanding caring as a form of commitment shows why a care‐based ethics must endorse moral partialism. The approach set out in this article is a significant improvement on theories that attempt to justify partialism by appealing to second‐order principles. It is also an improvement on other arguments for partialism made within the ethics of care, which justify partialism by appealing to the limitations on our knowledge of how to provide good care to unknown or distant others.

Of course, philosophers opposed to moral partialism might see my argument as providing them with a reason to reject care ethics, rather than with a reason to accept partialism. In closing, then, I wish to point out that my argument has implications beyond care ethics, and that it cannot simply be dismissed by those working in other ethical traditions.

First, virtue ethicists are increasingly coming to see caring as a centrally important virtue (for example, Reference McLaren, DesAutels and WaughMcLaren 2001; Reference HalwaniHalwani 2003; Reference SwantonSwanton 2003; Reference AnnasAnnas 2011). If this is the case, and given the argument of this article, it follows that virtue ethics is required to endorse some degree of moral partialism, in accordance with the importance that it attaches to caring. Of course, the extent to which virtue ethics will call for us to act on partialist considerations is debatable. But any approach to the virtues that accommodates caring will also need to accommodate some degree of partialism.

Of course, the ethics of care has often been discussed in terms of its relation to virtue ethics, so the extension of my argument to virtue‐ethical perspectives may be unsurprising. However, my argument may be still further extended to any ethical viewpoint that attempts to accommodate the moral importance of our commitments to other persons. For example, if a consequentialist theory permits us (perhaps at the level of second‐order rules) to form and act on interpersonal commitments, then it thereby also permits some degree of moral partialism. Indeed, the only way of avoiding moral partialism, as I see it, is to deny the moral appropriateness of interpersonal commitments altogether. This is a highly counterintuitive position to defend, however; indeed, it is much more counterintuitive than the more abstract claim that our moral duties are strictly impartial. By enabling us to see that impartialism requires the rejection of interpersonal commitments, then, my argument provides additional support for moral partialism even outside of the ethics of care.