1. The completion of care

Since Kittay builds her conception of the completion of care on foundations laid by Noddings and Tronto, I will begin by briefly exploring what they mean by “completion.” This background is important for understanding Kittay’s unique understanding of completion and setting the stage for my argument in the final section where I suggest that something closer to Noddings’s and Tronto’s conceptions provide a better standard for good care.

Noddings (2013, 4) states that care must be “somehow completed in the other” but never very clearly defines this concept. She does, however, give some indications of what she means by it. When discussing the perspective of the “cared-for,” she offers the following explanation of completion:

In this statement, Noddings associates completion with the perception or feeling by the cared-for that someone cares for them. She says care is completed when the cared-for perceives the “attitude” that someone has more-than-perfunctory regard for them and is not just going through the motions of attending to their physical or emotional needs.

Later, when discussing “The One-Caring,” Nodding supplements this description. “If the cared-for does not complete my caring by receiving and acknowledging it, I may examine myself and ask, ‘Do I really care?’” (Noddings 2013, 37). Noddings is mainly concerned here with negative situations where the care receiver rejects one’s care by saying something like “You don’t really care!” (2013, 38). Although caregivers might be tempted to brush off such accusations as uninformed or immature, Noddings argues they should take them seriously and use them to reflect on their caregiving. Even supposing the caregiver is providing a genuine benefit for the cared-for, something has gone awry in the caregiving relationship if the cared-for does not feel cared for. What that something is is the feeling of being cared-for as a distinct person whom the caregiver values in particular. Once again, Noddings thus associates completion with the cared-for’s recognition of the caregiver’s genuine concern for them.

Tronto (Reference Tronto1993) does not mention “completion” in defining the phases of care but instead identifies “care-receiving” as the final phase and “responsiveness” as the virtue associated with it (1993, 107–8, 134–6). Having attended to the caregiver’s needs and preferences in earlier phases of care, Tronto argues that in the final phase the caregiver gets down to the business of benefitting the person or thing they are caring for. They succeed—and thus “complete” care—when some noticeable improvement is achieved in the object of their care: “For example, the tuned piano sounds good again, the patient feels better, or the starving children seem healthier after being fed” (1993, 107). This is an objective, observable state. Because the caregiver may mistake what the care receiver needs or deliver it in less-than-optimal ways, Tronto adds that the virtue of responsiveness is important for realizing good care in this final phase. Responsiveness “suggests that we consider the other’s position as that other expresses it” (1993, 136, my emphasis). Notice here that Tronto does not say responsiveness involves doing what the other wants but only considering it in making judgments about what the caring thing to do is.

Kittay develops the concept of the completion of care more fully than Noddings and Tronto do but also differently from them.Footnote ³ Like Noddings and Tronto, she (2019, 184, 187) situates completion as the final phase or “endpoint” of normative or good care and argues that it involves positively affecting the well-being of others. “Care is an achievement,” she writes, “and ‘to care’ is a success verb” (2019, 187). However, whereas Noddings associates completion with the care receiver’s sense of being cared for and Tronto with successfully meeting needs in responsive ways, Kittay associates it with “recognition,” “taking up,” “active affirmation,” or “subjective endorsement” by the care receiver of the affective and material care provided for them (2019, 184–87, 198–200). “CARE requires that the object of the care respond in some way that results in the achievement of the act—that is, CARING requires that the cared-for ‘take up the action as care’” (2019, 187). Kittay’s understanding thus places more emphasis on subjective endorsement than Tronto does, alters the content of subjective endorsement from a general sense of being cared for, as Nodding argues, to affirmation of particular acts, and makes subjective endorsement a necessary condition of good care. Although the “general assumption” in thinking about caregiving is “that to be a recipient is to be passive,” Kittay argues, “this is false”: “There has to be an uptake on the other’s part if our action is to count as a benefit for the other” (2019, 186).

Kittay notes that the active endorsement of care by the care receiver can be either implicit or explicit (2019, 139). It does not require anything so obvious as gratitude. Yet it does involve something more than what Noddings and Tronto indicate is necessary for good care. It goes beyond satisfying objective needs in attentive and responsive ways, for example. Kittay (Reference Kittay2019, 186, n. 4) notes in this regard that Tronto offers “no discussion of the reception phase that could clarify how those persons or things that could not acknowledge or refuse care could be involved in care’s reception.” In fact, Tronto does not have much to say about subjective reception at all. Care is “completed” by her account when things and persons sound better and look healthier. While Tronto (Reference Tronto1993, 139) thinks the subjective input of care receivers is important to good care, she embeds this element in the final phase of care through the virtue of responsiveness. Care is good for Tronto, as Bennett (Reference Bennett2023, 497) notes, when it conforms to or embodies the “four phases” of care. Subjective endorsement is not a necessary feature of it.

Kittay argues that the nature of subjective uptake will vary depending on the “nature of the cared-for” (2019, 191). If the cared-for has no apparent subjectivity, uptake will be expressed through an objectively observable, physical process. If we pour water on a withered plant and it perks up, we can say it has taken up our care. Kittay’s notion of completion here seems close to Tronto’s. However, when the cared-for are subjects, “which is to say that they are capable of a subjective life,” Kittay argues, “caring aims to do more” than improve their objective welfare: “it aims to contribute to flourishing” (191). By contributing to “flourishing,” Kittay does not mean helping individuals to achieve an objective list of capabilities along the lines of Martha Nussbaum’s (Reference Nussbaum2000) capabilities approach. Rather, she means helping a person to do the things they care about (170). Flourishing is at root for her a subjective notion:

Because flourishing is at root a subjective notion for Kittay, she argues that we cannot promote it in purely objective ways. We must promote the things people themselves care about to promote their flourishing. This is why care requires subjective endorsement to be completed. The only way we can know if we have effectively promoted the flourishing of others is if they endorse what we have done for them.

Kittay notes that “there are objective measures by which we can determine that a person is not flourishing.” Subjective endorsement is necessary but not sufficient for good care (2019, 199). The role that these objective measures play in filling out the standard of good care is, however, limited and subordinate to subjective endorsement. Immediately after making her statement about the importance of objective measures to good care, for example, Kittay argues that suffering cannot be taken automatically as a sign that a person is not flourishing because some people consider suffering for others to be part of their understanding of a meaningful or flourishing life (199). She goes on to note that we should nonetheless be wary of the preferences of malnourished women who report being adequately fed even though their diets are worse than their husband’s. Even so, whether or not providing them access to more nutritional resources counts as caring still ultimately depends on their subjective endorsement (200, n. 33). We might aim at providing this good but whether our action will count as good care depends on their validation of it. Most generally, Kittay (Reference Kittay2019, 138–39, n. 6) writes that, although something “may objectively be (or commonly understood to be) a good” of a flourishing life, “it should also be good for the person concerned” if it is to be the goal of good care. In the final analysis, she thus argues that subject endorsement trumps objective measures of flourishing in the completion of care.

One of Kittay’s purposes in associating good care with subjective endorsement, as noted above, is to allay concerns about care’s paternalism. “Care is fundamentally opposed to a paternalistic imposition of a putative objective conception of a person’s good,” she writes (2019, 201, 208). It promotes the person’s subjective understanding of their good, not what some third-party might judge to be good for them. She (2019, 200–1, 207) helpfully contrasts her understanding of good care on this matter with Stephen Darwall’s theory of rational care. Darwall (Reference Darwall2002, 4) argues that good care involves successfully promoting a person’s good or welfare which “is constituted, not by what the person values, prefers, or wants (or should value), but by what one (perhaps she) should want insofar as one cares about her.” Like Kittay, Darwall argues good, or rational, care will aim to promote people’s flourishing, but not in the subjective sense she defends. Instead, he (2002, 75) associates flourishing with the enjoyment of “agent-neutral values such as aesthetic beauty, knowledge and understanding, and the worth of living beings” as well as more pedestrian welfare goods like health and safety.Footnote ⁴ Good care by his account means promoting goods that anyone who cares about a person (or sympathetically imagines the person to be someone they care about) would want for them from a “third-person perspective” (2002, 49).

Kittay (Reference Kittay2019, 201) rejects this objective definition as antithetical to good care: “On Darwall’s account, a carer has an objective standard for the other’s good” such that they are “justified in acting paternalistically and overriding the cared-for’s own idiosyncratic view of what is good and right for herself” at least in some cases. Her subjective endorsement principle reverses Darwell’s ordering of priorities by elevating the cared-for’s “own idiosyncratic view of what is good and right for herself” above an objective standard of what is good for people. Since the cared-for’s idiosyncratic preferences may be difficult to discern in some cases, particularly where a person cannot clearly communicate their preferences, Kittay discusses several difficult cases to show how it can guide action even when a person is unconscious or has limited cognitive capacities. If a person is temporarily unconscious, for example, we should care for them by doing what we think they will later subjectively endorse: “The action will be completed as such when the patient recovers” but only if they embrace what we have done to them (202). If they renounce what we did to them, or remain neutral toward it, we have not cared well for them. Similarly, Kittay argues that “good (enough) parenting often depends on such deferred reception” (203). Even if the child may not appreciate the parent’s action in the moment, parents can tell themselves: “One day she will appreciate what we did for her.” Care will not be completed until that day comes, which means parents cannot say they cared well for their children until their children subjectively endorse their parenting. If a child resents how they were cared for—fairly or not—the parents’ efforts to feed, cloth, educate, love, and encourage them will not count as good care no matter how objectively good their care might appear to have been from a third-person perspective (203–4).Footnote ⁵

Where a person with a severe cognitive disability rejects some vitally important form of care, such as taking a medication, Kittay suggests we should construct a counterfactual to discern their subjective endorsement: “If the cared-for could understand, then she would endorse my actions as CARE” (205). This informed-desire account of good care requires imagining what a person whose preferences are obscure would want if they were properly informed, cognitively and emotionally mature, and able to indicate their preferences to us. The counterfactual then provides the subjective endorsement of the care we provide for them.

Kittay discusses a couple of other cases but these examples should suffice to indicate how she thinks care might be completed by persons who might not be able immediately or explicitly to endorse it. Whenever any sort of subjective endorsement is possible, she is adamant that it is essential to the completion of good care.

2. Some doubts about Kittay’s subjective endorsement standard of good care

In this section, I explain why I think Kittay’s subjective endorsement standard is misguided, but before doing so, I want to reiterate the value of her argument. Kittay aims to clarify the standard of the final phase of good care to provide clearer guidance on how to care well for others and avoid worrisome forms of paternalism. Although I do not think her argument succeeds, it does bring to a head some important questions that have long haunted care ethics: How much of a care receiver’s preferences, or idiosyncratic views about what is right and good, should caregivers take into account when caring for them? How should they do so? Under what circumstances should they do so? What are the tradeoffs between associating good care with more objective or subjective goals? And what sort of caregiving standard is best suited to protecting against paternalistic abuse? Kittay’s argument brings all these questions to the forefront of the definition of care.

I have two main criticisms of Kittay’s subjective endorsement principle as a standard of good care: (1) it does not provide a reliable standard for distinguishing good care from bad; (2) it can justify a pernicious form of paternalism. To be clear, I do not mean to suggest that Kittay’s subjective endorsement principle merely has difficulty addressing some of the hard cases she discusses. My criticism is that the standard is inherently flawed.

In the first place, Kittay’s subjective endorsement standard is flawed because it cannot reliably and plausibly distinguish good care from bad. Kittay, it will be remembered, claims that subjective endorsement is necessary for good care. The problem is that people are not always reliable judges of good and bad care in their own cases. People-pleasers may thank someone for providing them the worst of services because it is their disposition to accept with gratitude whatever is done for them and they are too timid or conflict-averse ever to condemn or confront anyone. Because their subjective endorsement is effectively secured in advance for good or bad care, it is not a good indicator of either (unless we want to bite the bullet and say good care is whatever anyone does for them and they reflexively endorse). Grumps and curmudgeons, by contrast, may complain that nothing is ever good enough for them. Their judgments are also therefore not a good guide of good and bad care. People more generally are often not suitably informed about the range of different types of care available to them to be able to say with authority if the care they received was good or bad. This is one of the reasons why care theorists argue that care is not well allocated through free market competition. Because we often end up receiving care from only one source in a one-off manner, it is difficult to shop around for good care and know if the care one has chosen (e.g., from a doctor or childcare provider) is good or not. A patient whose surgeon has botched their operation may still praise the care they received because they have no sense of what better, or good, care might look like, particularly if their surgeon assures them that they were provided with the best possible care under the circumstances. Their subjective endorsement surely does not make their doctor’s objectively bad care good.

Here Kittay would likely want to bring in objective standards to condemn the surgeon’s care. Yet this move does not sit well with her principle of subjective endorsement. If a person subjectively endorses the care they have received, who are we to second guess them? Kittay might suggest that we should construct a counterfactual here to save the notion of subjective endorsement: if the patient had full information about good care and a fuller sense of the different types of care available to them, they presumably would not endorse what the surgeon did to them. As distinct from constructing a counterfactual to discern what someone who cannot give explicit consent might want, however, this counterfactual would override the explicit subjective endorsement of someone who can clearly communicate it. It would thus seem to invalidate the whole idea of subjective endorsement. A straightforward standard of objectively good care would seem sufficient and better here.

Kittay’s subjective endorsement standard of good care is also not a good guide for distinguishing good care from bad because people’s conceptions of good care are often shaped by the care they receive and the caregivers who provide it. People may be inclined to subjectively endorse whatever care they receive because it provides the main or only standard of good care they have. This problem is most obvious in the parent–child relationship but can also occur in more limited, short-term relationships where one party has more power (due to position, status, charisma, etc.) to influence the understanding of good care than the other. Doctors, for example, are often able to tell patients in effect what they should expect good care to look like. Unless patients are knowledgeable and assertive, they may just go along with their doctor’s definition. In the parent–child relationship, parents can even more deeply influence their children’s conceptions of good care. Children will assume in many cases that the care their parents give them is good care and if parents reinforce this idea by telling them their care is good (and even more if they isolate their children from examples of different or better caregiving), children will likely grow up to believe the care they received from their parents is good care. Just because children who are raised in cult-like environments endorse the care their parents gave them as good, however, does not make it such. For a relational theorist, Kittay gives remarkably little attention in Learning from my daughter to the ways in which people’s idiosyncratic ideas about good care are themselves shaped interpersonally. By associating good care with subjective endorsement, her theory can end up classifying as good care what people in authority want the people dependent on them to think it is.

Secondly, beyond just being misleading, Kittay’s subjective standard of good care is also more susceptible to certain forms of paternalistic abuse than the objective standard she rejects. This is clearest in her discussion of cases that are at the heart of her book: caring for children and adults with a cognitive disability.Footnote ⁶ In discussing her daughter Sesha’s resistance to taking some medications, Kittay notes that “if the medication works … I am CARING for her by any objective measure” (2019, 205). But objectively benefitting Sesha is not enough for normative care by Kittay’s account. She therefore suggests that we should construct a counterfactual to determine what Sesha would want if she were fully informed and had different cognitive or emotional capacities and were capable of verbal negotiations. As Kittay (205) recognizes, it can be dangerous to construct counterfactuals to determine the values and preferences of someone who cannot clearly and explicitly express them. Even so, she claims they are still better than objective standards because they at least attempt to discern what the care receiver would want if they were capable of making their wishes known.

In the final chapter of Learning from my daughter, however, Kittay effectually argues against this counterfactual approach to determining good care and endorses objective standards as a better guard against paternalism. She does so when discussing the case of Ashley X. Ashley X was a 6-year-old girl with severe cognitive disabilities when in 2002 her parents requested and were granted permission to have high doses of estrogen administered to her to arrest her development and to keep her forever small.Footnote ⁷ The parents also requested and were granted permission to have a hysterectomy performed on her and to have her breast buds removed. As Kittay acknowledges, the parents offered many reasons for thinking that Ashley would hypothetically consent to these procedures. They would mean, for example, that she could live forever with them rather than having at some point to move into an outside home with professional help. They would also reduce her susceptibility to sexual abuse and various cancers. Kittay takes “her parents at their word when they say that they did this for Ashley’s sake” (2019, 222). She nevertheless condemns these procedures not because Ashley hypothetically would have rejected them or because the future Ashley would not take them up—having had her physical development truncated and not knowing any other sort of life, one can imagine that she might endorse them—but because it runs afoul of objective notions of a flourishing human life. Kittay compares what Ashley’s parents did to her to pouring vinegar on a plant—an objective harm (2019, 223). She identifies “growth and sexual development” and “bodily integrity” as “crucial intrinsic goods” of human beings (234–35). She points to the disability community’s condemnation of these procedures as evidence that from an informed “third party” perspective this is not the sort of thing that one who sincerely cared for Ashley would do for her sake (228). She further asks:

In this passage, Kittay explicitly sets aside the subjective endorsement standard (i.e., “even if Ashley herself will never understand the growth attenuation as a slight”) to argue that the curtailment of “her right to grow and develop” was wrong because it infringed an objective feature of flourishing (i.e., having the opportunity to grow into one’s own person apart from Mom and Dad). Approaching the question of how best to care for Ashley by asking counterfactually what she would have wanted if she could express her preferences, by contrast, is what led her parents to convince themselves that they were doing the right thing by harnessing her development. Using counterfactuals and hypothetical consent, they formulated a notion of subjective endorsement that justified to their minds doing to Ashley what objectively from a welfare or capabilities perspective is appalling.

3. Rational and virtuous care

If the subjective endorsement principle is a misleading and possibly even dangerous standard for good care, what might we use instead? As stated above, I think something along the lines of Darwall’s rational care theory, amended to include elements from Noddings’s and Tronto’s theories, can effectively fill this slot. Objective standards can protect against the kind of abuse evident in the Ashley X case while intersubjective virtues like attentiveness, responsiveness, and respect can protect against the paternalism of objective standards. What is important is to keep a balance between the two rather than giving undue weight to either end of the spectrum (which is not to say that caregivers may not sometimes still legitimately act for objective goods over subjective preferences).

Darwall argues that we care well for a person when we successfully promote their welfare and flourishing in ways that people who sincerely care for them desire for their sake. At the heart of this conception of rational care is the notion of agent-neutrality. Agent-neutrality means eschewing idiosyncratic or agent-relative values and preferences of both the caregiver and care receiver when determining how to care for someone and instead doing “what anyone, perhaps she [the care receiver] herself, would rationally want for her … insofar as he or she cared about her” (Darwall Reference Darwall2002, 17, 83). “What benefits the cared for,” Darwall (Reference Darwall2002, 15) writes, “seems a good thing absolutely (agent-neutrally)” based on what is good for anyone and “not only good for him.” Darwall (Reference Darwall2002, 49) proposes this agent-neutral, or “third-person perspective,” to apply strictly to questions of what sort of care should be provided to people but not questions of whom we should care for or whether our affection for some particular persons can justify moral partiality toward them. It is strictly a matter of what kind of care to provide for a person for their sake.

Darwall concentrates on developing a metaethical theory of rational care and consequently does not say very much about its normative content. He does, however, identify goods like physical and mental health, beauty, and knowledge with it (2002, 5–6, 43–44, 50–51, 75). These are things that are good for human beings—or at least potentially good—given our bodily and mental constitution, regardless of what a particular person values and prefers.Footnote ⁸ Darwall’s shortlist of things that are good for human beings might be expanded to include other goods like adequate nutrition, decent shelter, bodily integrity, protection from harm, education, emotional support, social opportunities, etc., all of which are implied by the aims of rational care. These are all important elements of human welfare and flourishing and the sorts of things that one who sincerely cares for another wants them to have access to for their own sake. Darwall’s definition of rational care also helpfully provides a rough standard of the level at which and manner in which goods should be provided to people to be considered good care. Healthcare, for example, should be delivered at a level and in a way that a person who sincerely cares about a person would want for their sake.Footnote ⁹

Many care, feminist, disability, and other theorists are understandably skeptical about claims of neutral or objective goods but where the content of rational care is filled with basic goods like food, shelter, and bodily integrity, and the goal is said to be providing ready access to these goods rather than forcing them on people, and, as discussed below, access to these goods is provided in attentive, responsive, and respectful ways that take account of individuals’ preferences, these concerns appear less pressing. To be agent-neutral in the sense Darwall describes means nothing more (or less) than putting aside one’s own biases and trying to cut through the self-abnegation, uncertainty, or confusion of others to do what is really caring for them—i.e., supportive of their welfare and flourishing—in a given situation. It requires some measure of humility, critical self-reflection, and engagement with others to arrive at good care.

Darwall’s rational care standard avoids the main problems of Kittay’s subjective endorsement principle. It offers a clear standard for differentiating good from bad care that does not unduly weight the judgments of care receivers. From the perspective of rational care, it does not matter if a patient subjectively endorses the botched efforts of a surgeon to heal them. The surgeon’s care is bad if it does not meet the standard of what one who sincerely cares for the person would want for them—that is, a reasonable standard of good medical care. It likewise does not matter for Darwall’s rational care theory if a child who has been habituated to the emotional abuse of their parent comes to associate it with good care. Emotional abuse is hurtful and debilitating and thus not the sort of thing that one who sincerely wants to promote the welfare and flourishing of another should do to them for their sake.Footnote ¹⁰ Rational care likewise supports making children eat their vegetables, period, without the need to wait for their subjective endorsement to know if one’s care was good in this respect. In fairness, Kittay does say that care which is not objectively good is not fully normative even when it is subjectively endorsed, but as discussed above, her argument is ambiguous because she also leans heavily on subjective endorsement (i.e., a person’s idiosyncratic understanding of flourishing) to say what is objectively good care for them.

Darwall’s concept of rational care also provides stronger protections against paternalistic abuse than Kittay’s subjective endorsement principle. This might seem surprising since Kittay proposes her subjective endorsement precisely to avoid paternalism. As I have tried to demonstrate, however, when the principle is cashed out, it does not have the expected results. The Ashley X case in particular demonstrates the dangers that arise when caregivers orient their care around a person’s supposed or hypothetical preferences. One can talk oneself into all sorts of things that a person might value counterfactually. Notably, this danger is not exclusive to caring for people with disabilities but also applies to care for non-disabled children and other people whose subjective preferences might not be fully formed or explicit. Rational care provides a check on this sort of counterfactual projection by limiting good care to the pursuit of objective goods of welfare and flourishing. The fatal moral flaw in the Ashley X case was making an exception for how she was cared for based on her supposed idiosyncratic preferences. Rational care theory suggests that Ashley’s parents and doctors should have wanted for her what anyone would want for someone for whom they sincerely care. This is to grow up and gain some measure of independence from their parents or caregivers if possible and to experience life as a unique individual based on their unique capabilities. This standard for good care is crystal clear. If people do not find it acceptable to stunt the development of non-disabled children to keep them safe at home with Mom and Dad, then we should not be doing it, or allowing it to be done, to disabled children. While protection against sexual abuse is obviously an important concern for all people (disabled or not), there are better ways to secure it (for disabled and non-disabled people alike) than incapacitation or imprisonment.Footnote ¹¹

Although Darwall’s agent-neutral theory of good care better protects against agent-relative paternalism of the sort exhibited in the Ashley X case, Kittay’s subjective endorsement standard might be thought to provide better protection against run-of-the-mill paternalism of the sort where a caregiver promotes some objective interest of a person over their preferences. The seeming superiority of Kittay’s standard is largely illusory, however, for reasons spelled out in the last section. People may subjectively endorse the care they receive for all sorts of reasons, including that they have been habituated and indoctrinated to endorse it. Subjective endorsement does not therefore address one of the main concerns about paternalism—that people will be subject to the will of another and cared for in ways that are contrary to their interests. It allows consent to provide cover for what may already be thoroughly paternalistic or dominating relationships. Darwall’s theory, by contrast, provides an objective check on what can be considered good care—and hence what a caregiver can do to others—regardless of the care receiver’s subjective endorsement.

Darwall’s rational care theory further provides protections—or can be extended to do so—against the paternalism of objective standards. In most cases, it should first be noted, Darwall’s rational care theory does not justify paternalism at all. It directs caregivers to promote the sort of agent-neutral goods that most people want most of the time. Good caregivers, by Darwall’s definition, do what people who truly care about a person, including the person herself, want someone to do for her, e.g., help her find adequate food, clothing, education, emotional support, etc. As long as the person wants these things, paternalism is not an issue. The rub comes only when someone wants something that anyone—not just their particular caregiver—would not recognize as consistent with their good care (e.g., a child wants to eat a giant bag of candy for dinner every evening). Then the caregiver, if they are to care well for the person, will act to promote agent-neutral goods, such as providing nutritious food and promoting healthy habits, rather than acting for their preferences. Even then, Darwall argues, rational care should not always prevail over agent-relative values. In many cases, he argues, respect should have precedence over the objective goods of rational care. When we respect a person, Darwall (Reference Darwall2002, 14–15) argues, we recognize their capacity for free agency and the importance of their agent-relative values and preferences to them. We show people respect by prioritizing these agent-relative preferences and beliefs over objective care. Respect thus provides a check on the potential paternalism of rational care. Depending on the circumstances, we may forego promoting the objective goods of a person and instead respect their idiosyncratic preferences and beliefs. “We may rightly think that unhealthy habits are harmful for someone,” Darwall writes (2002, 14), “but think as well that respect tells against exerting undue pressure to induce her to change.” Our understanding of good care does not change in these cases, which is what Kittay’s argument suggests, but our understanding of what we should do does. We should forego care for respect at least where circumstances (discussed below) warrant it.

Darwall sharply distinguishes care from respect, but Kittay (Reference Kittay2019, 210–13) and other care theorists (Engster Reference Engster2007, 31) include it as one of the virtues of care. Good care involves, by these accounts, recognizing the dignity of others and treating them as equals regardless of their needs or circumstances and recognizing their agency in determining how to act toward them.Footnote ¹² Kittay elevates this second sense of respect nearly to the equivalent of care itself: we care for others by respecting (in the sense of caring about) their idiosyncratic preferences and beliefs. A weaker sense of respect—treating it as one virtue of good care as opposed synonymous with it—makes room for objective standards of good care. Good care, from this perspective, involves knowing when we should promote the objective goods of people over and above their stated preferences and when we should hold back our desire to promote our understanding of a person’s objective good and instead honor their preferences. The difference from Darwall’s argument is just that he treats respect as a limiting condition on care that exists outside of it, whereas I am suggesting that respect should be understood as one expression of care. The inclusion of respect in a definition of good care makes sense if promoting flourishing is one of its goals. If we want to promote the flourishing of others, we sometimes need to hold back our desire to protect or shelter them so they can live their lives according to their own idiosyncratic preferences and beliefs. Giving a person freedom to pursue their preferences and dreams (e.g., going mountain climbing even though it is dangerous) is from this perspective not antithetical to good care or a limiting condition on it, as Darwall argues, but an expression of good care. It is also an inherent check on paternalism that exists within the ideal of good care itself.

Darwall highlights some of the circumstances that can help us to decide when we should promote the objective goods of care and when we should respect a person’s values in caring relationships. While we should insist that a young child “eat his broccoli” and abide by the rules we set to protect and support him, Darwall writes (2002, 15): “When, however, the child matures into a competent agent, then his will and desires do acquire independent weight.” For a parent to presume to tell a competent adult child what to eat and how to act “is paternalism in the pejorative sense.” Similarly, even if it is clear to us that the health of a person would be promoted by giving them some medication, if they are competent and do not want to take the medication for some good reasons, we should respect their decision “even though concern for her and her welfare tugs in a different direction” (Darwall Reference Darwall2002, 43–44). Formalizing these conditions, we might say that in caring for others we should prioritize respect to the degree that they are mature and competent (age being a rough approximation of these criteria) and able to understand the consequences of their actions (or of our acting according to their preferences) for their welfare. Where these conditions apply, good care will align closely with respect though, as Darwall notes, we may still feel the tug of wanting to encourage people to act for their welfare.

Darwall’s rational care theory can be further buffered against paternalism by including within it the other main virtues of care: attentiveness and responsiveness (Engster Reference Engster2007, 30–31; Noddings Reference Noddings2002, 14–19; Tronto Reference Tronto1993, 127–37). Because Darwall draws a stark contrast between care and respect, he does not consider the role that agent-relative preferences might play within an agent-neutral account of good care. Agent-relative preferences are invaluable, however, not only for limiting the scope of care where agency respect is due to a person but also for fitting agent-neutral goods to the specific features of different individuals’ lives. In this regard, they seem to be implied by Darwall’s concept of rational care. One who sincerely cares for someone would surely want to attend and respond to their preferences, values, and circumstances in attempting to meet their agent-neutral needs and interests. They otherwise will not convey to them that they care for them for their sake and may not even provide them with the goods they need for their welfare and flourishing. It is all well and good, for example, to provide the agent-neutral good of food to hungry people but if we send beef to vegetarians or pork to individuals whose religious beliefs forbid them from eating it, we will not successfully care for them. Likewise, if we try to force a love of music (or beauty) on someone who is more inclined to find joy in playing sports, we might impede their flourishing rather than promote it.

Including attentiveness and responsiveness along with respect in Darwall’s definition of good care greatly limits its potential paternalism (Tronto Reference Tronto1993, 145–46). A good deal of paternalism that exists in care arises from well-intentioned efforts to meet real needs or promote real goods of people (e.g., a love of music or baseball) but in ways not suited to their preferences—e.g., a father wants his son to be a boxer when the boy wants to be a dancer.Footnote ¹³ Allowing individuals to have input on the particular nature of the needs and goods that support their welfare and flourishing significantly reduces this danger. The emphasis remains on promoting agent-neutral goods like good nutrition, health, knowledge, sports, and beauty but with input from individuals about how these goods, and which of these goods, should be realized in their cases. Parents will still require their children to eat their vegetables and other healthy foods under this ideal of good care but will make an effort to determine which sorts of vegetables (or perhaps fruits) their children like best and provide them with these nutritious foods rather than dogmatically forcing them “to eat their spinach.”

Even when amended in this way, Darwall’s definition of care ethics still justifies a kind of paternalism in a limited number of cases. When individuals do not appear competent or capable of understanding the consequences of their actions or seem to be acting impulsively or irrationally, a caregiver might legitimately decide to promote objective goods for them like good nutrition, health, and social connection possibly over their preferences. Although the caregiver acts in these cases paternalistically, it may be said in their defense that they at least act according to an agent-neutral form of paternalism rather than an agent-relative one. Their paternalism is not based on idiosyncratic ideas about what is good for a person but objective ideas about what is good for human beings. It therefore lacks the perniciously dominating quality of agent-relative paternalism. Individuals may still resent being cared for in a manner that does not align with their subjective preferences but they will at least be able to understand the logic and purpose behind the caregiver’s actions and thereby know that they mean to care for them. Their sense of being arbitrarily subjected to the will of another will in this way be mitigated.

Admitting that good care can sometimes involve some measure of paternalism might seem a flaw of this theory but I consider it a strength. I do not think a plausible theory of good care can escape the danger of paternalism altogether. To care for someone is to wish to promote some good for them. We can at best check the paternalism inherent in our care by orienting it around goods that all people generally want and incorporating virtues of attentiveness, responsiveness, and respect into it. But we should not pretend that we can construct a non-paternalistic theory of good care theory by stipulating, for example, that care is only good when care receivers subjectively endorse it. Such an understanding of good care invites problems as bad as the ones it is supposed to resolve. Rather than trying to evade paternalism altogether, we should work constantly to care for others in ways that promote their good and take account of their preferences. We will inevitably sometimes get the balance wrong but at least if we are aware of the danger we can try to self-correct and improve our care. What we should not do is pretend that the care receiver’s explicit or implicit endorsement of our care can relieve us of this worry.