Anscombe's Philosophy of Action

What then do we mean by “action” and “practice”? Care ethicists have not fully fleshed out an account of what actions in general are or how care fits into such an account. But there are numerous theories of action on which they might lean. The one I shall defend is Anscombean. One of the most pressing ethical issues in Elizabeth Anscombe's eyes, which she addresses in both “Modern Moral Philosophy” and Intention, is what is sometimes known as the “problem of relevant descriptions” (Anscombe Reference Anscombe1958; Reference Anscombe1963). It is foundational to Anscombe's view, Onora O'Neill tells us, that: “action is propositional. Acts fall under many descriptions; act-descriptions may have many instances. Although individual acts—act tokens—are events in the world, we both think about action and act under certain descriptions. We consent to action as described in certain ways, but not others” (O'Neill Reference O'Neill2004, 305, original emphasis). But this gives rise to problems in ethics. Anscombe argues that Kant's “rule about universalizable maxims is useless without stipulations as to what shall count as a relevant description of an action with a view to constructing a maxim about it” (Anscombe Reference Anscombe1958, 2), and like Kant, Mill “fails to realize the necessity for stipulation of relevant descriptions, if his theory is to have content” (3). Both Kant's and Mill's accounts therefore succumb to what Anscombe regards as a fatal flaw: “Any act-token will fall under many true descriptions, hence also under many possible principles of action. Absent a proper philosophy of psychology, how can we tell which act description is relevant for moral assessment?” (O'Neill Reference O'Neill2004, 306).

Anscombe goes on to develop a means by which act-types can be individuated. As Candace Vogler summarizes it, on Anscombe's account, “an act-type is just any isolable, determinately describable aspect of an exterior act done on purpose, such that the agent can answer a question about what he's doing couched in terms of that description” (Vogler Reference Vogler, Hoffman, Müller and Perkams2013, 245). Swinging one's arm can be a token of numerous act-types, depending on whether one is hacking wood, dancing, gesturing, or bowling a ball. Each description is isolable, able to be disentangled from other actions of which it might be an element or with which it might form a sequence (or one complex action). Each description also makes for an intelligible response to the question “what are you doing?” And each of these descriptions represents an aspect of an exterior act that is performed on purpose (unlike a reflex) and is determinately describable, such that it can be explained in detail. In the case of care, we might say that if one were to ask pedestrians picking up litter what they were doing, “recycling cans” would be a perfectly intelligible response, as would “taking care of the environment” or “impressing a love interest,” all of which can refer to the very same act-token.

A subset of act-types is intentional. Anscombe tells us that intentional actions “are the actions to which a certain sense of the question “Why?” is given application; the sense is of course that in which the answer, if positive, gives a reason for acting” (Anscombe Reference Anscombe1963, 9; and note that Anscombe does not think that the question needs an answer, just that it be applicable). Anscombe then spends a large part of Intention describing what exactly is going on when one asks “why?” and what is going on when the agent answers it. She offers the following summary of her conclusions:

Anscombe is cognizant of the variety of reasons we can offer in explanation of an intentional action, and she deliberately leaves room for this.

Anscombe does not think that applicable descriptions of particular actions are all isolated from one another. Rather, she suggests that intentional actions can sometimes be described with increasing abstraction or remoteness by what she calls an “A-B-C-D order.” Anscombe writes that a man “moving his arm up and down with his finger round the pump handle is, in these circumstances, operating the pump; and, in these circumstances, it is replenishing the house water-supply; and, in these circumstances, it is poisoning the household” (46). In performing each of these actions, Anscombe's man also performs the next—Anscombe is making an identity claim about these distinct act-types. B, C, and D also give some explanation of why I am A-ing. So it makes sense, on an Anscombean view, to say both that S bandages a wound in order to heal a patient, and that in bandaging the wound she is healing the patient. Anscombe also does not think that in order for an action to qualify as a particular act-type one must succeed in bringing about its intended consequences (§23). On Anscombe's view, I A by B-ing whether or not I succeed at A-ing or B-ing—Anscombe's man can be described as poisoning the household even if, unbeknownst to him, some filtration mechanism had purified the water before it could be pumped into the house's cistern.

Care Ethical Philosophies of Action

Care ethicists are not committed to this account of action, but here I shall argue that what care ethicists have given us of a care ethical theory of action is largely consistent with Anscombe's, and that Anscombe's theory of action provides a suitable conceptual scaffold for an account of caring actions. Establishing consistency is not a straightforward comparative project, however, since care ethicists have not devoted much time to developing a theory of what actions are. But it remains possible. Though they have not yet advanced a comprehensive theory of action, care ethicists have, in the course of developing their theory of practice, offered several means of distinguishing actions/activities from practices, and in doing so have made some contributions to an account of caring action. Virginia Held, for instance, maintains that practices are one species in a broader genus, activity, into which care falls. In The Ethics of Care, she writes: “An activity must be purposive to count as work or labor, but it need not incorporate any values, even efficiency, in the doing of it. Chopping at a tree, however clumsily, to fell it, could be work. But when it does incorporate such values as doing so effectively, it becomes the practice of woodcutting” (Held Reference Held2006, 37; see also Held Reference Held2014, 111–12). Though the two concepts may come apart, Held does not seem to distinguish between action and activity. Here I take her to be treating the two as synonymous, in part because she draws on examples similar to those drawn on by philosophers of action (including Anscombe) to illustrate the phenomenon she is describing (see also, for example, Held Reference Held1990). Held understands both labor and practice as subcategories of action. For Held, work and labor are purposive activities, and a practice is a value-laden, purposive activity.

Of course, there are questions as to whether Held's “purposiveness” is synonymous with Anscombe's “intentional,” and there is almost certainly not enough in Held's corpus to make an incontrovertible claim either way. But to say that an action was performed purposively suggests that there is some answer to the question “why?” because acting toward some purpose seems to involve acting for the sake of bringing about some desired or otherwise sought-after end (it might even entail this, since one homonym of “purpose” is synonymous with “aim” or “end”). Held's example of chopping at a tree with the aim of felling it (to wit, with an answer to the question “why?”) is illuminating here, and her subsequent definition of practices as value-laden is certainly not inconsistent with Anscombe's theory of action—an Anscombean would simply interpret felling the tree as the reason for chopping at it. Her understandings of “work” or “labor” are also obviously especially broad—perhaps so broad as to be coextensive with intentional action, though they need not be so broad in order to be compatible with Anscombe's account thereof. What fragments we have of Held's theory of action do not, consequently, rule out Anscombe's philosophy of action as a philosophical underpinning for a theory of caring actions.

Held is not the only care theorist to have made contributions to a theory of action. Joan Tronto also sees the practice of care as a set of activities coupled with a particular “habit of mind” (Tronto Reference Tronto1993, 127), which Tronto tells us includes purposes (Tronto Reference Tronto2010, 161) but presumably also includes such things as values, though with the caveat that those activities are ongoing. Again, I see nothing in Tronto's discussion of practices to suggest that she is hostile toward an Anscombean philosophy of action. We have already seen how Anscombe's philosophy of action can accommodate purposiveness conditions. Moreover, like Anscombe, Tronto thinks that practices can be nested according to their proximate ends, and like Anscombe she suggests that particular practices, defined by their proximate ends, can be attached to a variety of distant ends (Tronto Reference Tronto2013, 21).

Sara Ruddick is another care ethicist who has contributed to a care ethical theory of caring actions. In Maternal Thinking, she writes: “Practices are collective human activities distinguished by the aims that identify them and by the consequent demands made on practitioners committed to those aims. The aims or goals that define a practice are so central or “constitutive” that in the absence of the goal you would not have that practice” (Ruddick Reference Ruddick1995, 13–14). Ruddick views practices as a subcategory of activities undertaken collectively, and distinguished teleologically (see Sevenhuijsen Reference Sevenhuijsen and Savage1998, 21 for a similar definition). Anscombe's Aristotelian predilections are mirrored in Ruddick's teleological method of delineating practices, and the Anscombean can make sense both of Ruddick's additional proviso that practices are defined in part by the demands their telê impose on the practitioner, and her proviso that practices are collective undertakings, simply by treating practices as a subcategory of actions with additional features. That practices necessitate particular act-types—that mothering necessitates tending to wounds, for instance—coheres perfectly well with the Anscombean claim that actions admit of increasing levels of abstraction in A-B-C-D order, just as it would make sense on Anscombe's view to say that the act of killing the inhabitants of a household necessitates that one inflicts some harm upon them. So Ruddick's (fragmentary) philosophy of action and Anscombe's do not seem to be inconsistent. Indeed, the synergies suggest that the two would mesh well.

The Heterogeneity of Care and Eudaimonistic Needs

“Although ‘need’ has been the central concept of the ethics of care from the very beginning,” Gheaus writes, “it has so far remained relatively under theorized. Various authors on care have intuitively used it to refer to most of what we usually call ‘needs’ in common language, that is biological as well as emotional needs” (Gheaus Reference Gheaus2010, 621; see also Clark Miller Reference Clark Miller2012, chap. 1). Many of care ethics’ earliest and most renowned proponents define need ostensively, pointing to cases that any plausible theory of need will admit. Needs for food and shelter have thus gone unargued in the literature, as have needs for mothering, education, and healthcare. Ruth Groenhout exemplifies the strategy, listing mothering, “the active tending of infants and children,” tending to the elderly, and friendship (Groenhout Reference Groenhout2004, 28–29; see also, for example, Tronto Reference Tronto1993; Held Reference Held2006). Peta Bowden and Grace Clement are also noteworthy, as we have seen, for refusing to define care and opting instead for the use of paradigmatic examples like these (Clement Reference Clement1996; Bowden Reference Bowden1997). Engster does not describe them as such, but he offers a similar list of what we might take to be paradigmatic cases of care (Engster Reference Engster2005, 55–56). Engster relies on this strategy repeatedly in fact, elsewhere offering the following list: “feeding the hungry, providing medication to the sick, teaching a child to walk or talk, sheltering or clothing someone, and helping a person to regain basic functioning after an accident” (Engster Reference Engster2006, 522). This approach has its advantages. As Tronto puts it, “the question of trying to define and to specify ‘needs’ is a difficult problem, both politically and philosophically” (Tronto Reference Tronto2013, 162), and thus it is reasonable to “articulate another approach that avoids ‘needs’ and focuses instead on ‘basic human capabilities’” (162; cf. Tronto Reference Tronto1995). The care ethical strategy of using ostensive definitions in lieu of formal definitions supplying necessary and sufficient conditions is also reasonable in light of this.

Yet Gheaus offered the remarks above a decade ago, and care ethicists have done work in the intervening years to develop a philosophy of need. Clark Miller offers one account of care according to which caring actions meet “fundamental needs,” defined as “needs that threaten agency in the sense that if they are not met, the serious harm of compromised agency will result” (Clark Miller Reference Clark Miller2012, 17). Fundamental needs, in other words, “must be met in order to establish, maintain, or restore agency” (17). Clark Miller has a relatively permissive understanding of agency, moreover, which includes not only the familiar set of needs such as freedom from certain sorts of interference or coercion, but also certain emotional and relational needs. This set of needs is thus fairly expansive, and Clark Miller's account of needs yields one elegant solution to the problem of care's heterogeneity: define care by reference to need, but define need broadly. Yet even this more permissive account of care seems vulnerable to a number of counterexamples. Consider the following case:

It is perhaps a little jarring to suggest that A needs the cake, since she can complete her work without it and will not be much worse off. It is less odd to say that A desires it. But it still seems natural to describe B's actions as caring. Consider also another counterexample:

Once again, we have a case where it seems natural to describe an agent as a caregiver, and to describe the actions they perform as caring. But the aid C's parents give her does not seem to count as meeting a need—C is perfectly capable of leading a rich and fulfilling life if she attends a mid-tier school. Nevertheless, we seem to think that there is something caring about C's parents acting in her interests (that is, benefiting her or giving her some advantage even if it is not needed).Footnote ²

On the view I have just presented, care represents a distinct and unitary act-type because its purpose/intention is to meet needs. But if the counterexamples above are convincing, then the definition of need we have just offered seems to run up against everyday locution. Care, as we use the term in ordinary language, also aims sometimes at satisfying desires or furthering interests, and one could surely come up with cases where one cares by accommodating preferences. Accounts like Clark Miller's identify a subset of objects that care might aim at—indeed, they are some of the most morally demanding, and Clark Miller is right to focus on them—but care's scope is broader than accounts like these admit. Much care consists not in meeting fundamental needs but in bringing joy, doing small favors, and otherwise making slight improvements in the lives of those for whom we care. Desires, interests, and preferences may not be as morally demanding as fundamental needs, but it does not follow either that they are not objects of care or that they are morally inert, and this is not how ordinary language would have us define care. And if this is the case, then we still appear to have several distinct act-types with several distinct proximate ends. In other words, we do indeed equivocate in describing all of these actions as “caring.” One option here is to simply bite the bullet and accept that the care ethical lexicon diverges from ordinary moral language, and to give the actions in the counterexamples above a different descriptor. In this section, however, I would like to defend an account of need that enables us to explain the unity of caring actions despite the apparent heterogeneity of the items in care's ambit.

Anscombe advances a theory of “Aristotelian necessity” according to which a statement that some person P needs x is to say that she will not flourish unless she has x. On this view, “to say that [an organism] needs that environment is not to say, for example, that you want it to have that environment, but that it won't flourish unless it has it” (Anscombe Reference Anscombe1958, 7). Anscombe's Aristotelian-Thomistic account of flourishing is one of many we might draw on, but I shall not commit to one specific concept of flourishing here. The usefulness of the concept can be shown by reference to elements of flourishing that are mainstream.Footnote ³ Eudaimonistic understandings of need can encompass all of the paradigmatic cases of care identified by care ethicists. Children cannot flourish unless they have a loving parent who takes care of them. The infirm cannot flourish unless they have medical professionals like nurses and doctors to heal them. Because human beings cannot flourish outside of loving relationships, and because they are corporeal beings who can be injured or harmed, they need caregivers to help them grow and recover from injury. This is not to say that needs only make sense within this eudaimonistic framework. There is no need here to suggest that a child's claim to need a toy is nonsensical, since they may well need it in order to maximally enjoy their afternoon. But before we regard indulging the child as caring for them, we ask eudaimonistic questions—is this child leading a good life? Have they been spoiled? Is the money better spent on food or warm clothing? What's more, it is only once we answer such questions that the need exerts normative force. Whether that need is morally compelling—that is, whether that needs-claim accurately represents some state of affairs that we ought to factor into our moral deliberation—turns on whether the toy is necessary for flourishing. So eudaimonistic accounts of caring actions appear to be particularly useful as foundations for ethical theorizing.

Concepts of flourishing thus cover what care ethicists would like care to cover, but they also solve the problem of heterogeneity. This conception of need broadens care's scope significantly. In particular, it is no longer clear that care cannot aim to satisfy needs and desires, interests, and preferences. One could now argue, quite reasonably, that the flourishing life involves at least some satisfaction of desires, some aid in the pursuit of one's interests, and some accommodation of one's preferences. Indeed, this is a standard view among philosophers who discuss flourishing. Julia Annas, taking flourishing to be synonymous with happiness, writes: “Happiness must involve our enjoying the good things in life. . . . [We] demand that happiness involve what we normally take to be enjoyment of life's advantages, and to the extent that we do, we will make happiness dependent on external goods” (Annas Reference Annas1993, 365). If this is the case, then one can care for others by satisfying their desires. But notice that because the approach here includes desires only insofar as desire-satisfaction is an element of the flourishing life, only certain kinds of desires are morally salient, and they ought only to be met in certain circumstances. Because the claim is that meeting desires is an aspect of care only because desire-satisfaction is one aspect of eudaimonia, one does not care if one satisfies wicked desires. The desire to torture puppies would never feature in a flourishing life (and should certainly never be satisfied in one), and one therefore does not care for others in enabling them to do so.

This, however, seems like a reasonable limitation on the claim that care involves desire-satisfaction. The sorts of desires we would hope to accommodate in allowing for desire-satisfaction include such things as the desire to catch up with an old friend, to make one's parents proud, and so forth. We are surely not going to admit that care can include satisfying desires to get revenge on a rival, to murder one's elderly parents for the sake of an inheritance, and so on. Good desires may not always need to be satisfied either, however, and here again the ancients can contribute to care theory. On one view popular in ancient Greece, to say that a person is flourishing is to make an evaluation of her life as a whole (Annas Reference Annas1993). Yet it is clear that particular desires, even when they are such as a temperate, honorable, classically virtuous person would have, are not typically serious enough to affect an evaluation of one's life as a whole. If Jane wants a specific book and never receives it, we do not think she is much worse off. But it does seem that the assessment changes if we say that Jane is an avid reader and never receives any of the books she would most like to read. Thus care may not require us to satisfy every good desire we encounter—we do not fail to care if we do not satisfy every good desire we are confronted with—but it does seem to require us to meet desires sometimes. So it appears that construing flourishing as care's end enables us to treat desire-satisfaction as an aspect of care without (a) equivocating, (b) endorsing the satisfaction of all types of desires, or (c) claiming that good desires must always be met.

Consequences and Good Care

Noddings once provocatively claimed that “care theory is consequentialist” (Noddings Reference Noddings2002b, 30). In doing so, she did not intend to align care ethics with thinkers like Bentham, Mill, and Sidgwick. Rather, she meant to emphasize the import of caring successfully. For Noddings, care ethics “demands to know whether relations of care have in fact been established, maintained, or enhanced, and by extension it counsels us to consider effects on the whole web or network of care” (30). Thus: “We cannot reasonably leave the description of caring at the purely phenomenological level. We want some sort of positive result for the cared-for, perhaps growth, as Mayeroff suggests, or happiness and reasonable comfort, as Butler suggests, or at least freedom from unnecessary suffering” (41). Care is not care, to Noddings, unless it actually meets needs. And Noddings is not alone in defending a view like this. Kittay too sees care as necessarily successful, writing: “I may intend to care for a parched plant by watering it, but if unbeknownst to me, the glass of clear liquid I pour in the plant contains vinegar, not water, most would agree that, despite my good intentions, I have failed to care for it” (Kittay Reference Kittay2011, 612).Footnote ⁴ If caring actions are always successful, then it follows that good caring actions will always succeed too. These statements about the nature of care sit in opposition to the Anscombean theory of care I have laid out in the preceding sections. According to the view I have defended, actions are caring if their proximate end is to meet eudaimonistic needs. This, I have said, does not entail anything about whether that action succeeds in bringing about its end. A nurse cares for a patient by administering medication whether or not that patient recovers; a parent cares for a child by helping her with her homework whether or not that child learns.

What is not in doubt, however, is that care theorists, like other ethicists, must make sense of consequences in their account of good care. On the one hand, the emphasis here on consequences reinforces what seems to be common sense. As John Rawls notes in A Theory of Justice, “all ethical doctrines worth our attention take consequences into account in judging rightness. One that did not would simply be irrational, crazy” (Rawls Reference Rawls1971, 30). Success criteria also serve certain discursive ends. For instance, they highlight certain important and hitherto underexplored categories of care, including, in particular, wise care. An experienced caregiver who excels at means–ends reasoning will be better able to care well. A wise caregiver in Kittay's scenario will be cognizant, for example, of the fact that she was cooking with vinegar earlier, and left a glass of it on the table near the plant. She may think first to smell or taste the liquid in the glass to see whether it is indeed water. A less wise caregiver may not be so thoughtful or perceptive, and thus be more culpable for the same failure to meet needs. In focusing on the outcomes of our efforts to care, we are led to ask whether there is anything the caregiver could have done better next time. Taking the actual satisfaction of needs as an end requires in part that we dissect cases of unsuccessful “care” with the aim of doing better in future. So the motivation for endorsing a success criterion is a reasonable one.

But I want to suggest here that consequences be left to evaluations of caring actions rather than descriptions of caring actions like those I have been discussing thus far. Success criteria are open to counterexamples:

Collins is right to think that morally praiseworthy care is sometimes unsuccessful. Caregivers who never commit any misdeeds may still have their efforts thwarted by luck. Failures to meet needs certainly represent a sort of disvalue, but that is not always the sort of moral value that ethicists usually attribute to actions. Collins thus seems right to suggest that an account of care that attaches value only to successful care is inadequate as the foundation for a normative theory.Footnote ⁵

Collins's counterexample also raises a descriptive point. It is not obvious that cases like Collins's, where I fail through no fault of my own to meet needs, are failures to care. At least as far as my own intuitions are concerned, it is jarring to describe one scenario's actions as caring and the other's as noncaring (or perhaps uncaring) given that they are identical except for the contributions of moral luck. The leap from the intuitive evaluative significance of consequences to the descriptive necessity of success in care is not one that can rest on common sense, since there is surely no consensus here. But given that we want the concept to act as a foundation for a moral theory, Collins's counterargument seems to render accounts that eschew all-encompassing success criteria preferable to those that include them. A third, related danger in including a success criterion for care and then using that concept as the foundation for a normative ethic is that it may yield an ethic antithetical to feminist aims. What such an account seems insensitive to is the extent to which successfully meeting needs comes down to the resources one has at hand. In many parts of the world, meeting needs requires material goods and the ability to purchase them: transportation, medical bills, and education often incur some (typically financial) cost. If an action must be successful to qualify as caring, and caring actions are morally praiseworthy, then a great deal of care-related praise will be reserved for the well-off because they are well-off. Care ethicists ought to prefer accounts that afford moral praise for acting as one ought while simultaneously affording (a different sort of) value to successfully meeting needs.Footnote ⁶ Successful care is a category of care we wish to retain, but it is one of several we ought to leave conceptual space for, including care that is nonculpably unsuccessful.Footnote ⁷ We ought to prefer accounts that make a distinction between success and failure within their concept(s) of care, instead of exporting failure to some other moral concept that is not the ethic's keystone moral concept.

These considerations seem to point us in the direction of those accounts that distinguish between something like “good” and “ideal” caring actions. Into the latter category falls care that actually meets needs, and into the former, actions that are marred by misfortune and yet still are caring. The latter takes into account all good-making features of caring actions, including both the agential and the nonagential. The former leaves room for action that is not successful, but not through the caregiver's own fault. We ought to hope that all care is ideal, and it is a pity when care is, for whatever reason, good but not ideal, but good care is still “good” in that the caregiver acted as they ought. Care ethicists would not be the first to distinguish between actions that are morally good but fail to achieve their ends, either. Virtue ethicists since antiquity have argued for a distinction between actions that are morally praiseworthy but not successful (Aristotle Reference Aristotle and Crisp2004), and even utilitarians, according to whom an action is morally right just if it maximizes utility, have come to distinguish between actual and expected utility in their moral assessments (Brandt Reference Brandt1959; Jackson Reference Jackson1991). Care ethicists might therefore choose to borrow this distinction from ethicists in other camps.

Caring Practices

Care ethicists, we have seen, often view caring practices as a subset of caring actions. In this section I explore the usefulness of this category for evaluations of caring actions. Whether and how care qualifies as a practice depends on which account of practice one buys into. In contrast to approaches like Held's, Tronto's, and Ruddick's, Alasdair MacIntyre defines a practice as:

This is a far higher standard than Held's or Tronto's, and MacIntyre has thereby excluded a number of activities that we might intuitively call practices (see, for example, Brewer Reference Brewer1997; Dunne Reference Dunne2003).Footnote ⁸ Rather than attempting to show that care qualifies as a sort of practice using one particular theory of practice or another, my strategy will be to argue that theories of practice remain analytically useful by appealing to certain widely accepted elements thereof.

Practice's social dimensions are one of the more interesting of these. The fact that a subset of caring actions is social adds several layers of complexity to the account of good caring actions. As Tronto points out, care's social nature introduces a set of external power dynamics to caring relations (Tronto Reference Tronto2010; Reference Tronto2013). One thought underpinning her work, and the work of care ethicists like Held and Fiona Robinson, is that caring relations are subject to authorities beyond the caregiver–recipient relation (Robinson Reference Robinson1999; Held Reference Held2006; Reference Held2011; Robinson Reference Robinson2011). Caring relations are caught up in the power dynamics of the wider family, community, profession, state, and even the international community. Most obviously, these relations bear upon the success or failure of our efforts to care. Care is subject to a variety of norms, standards, and laws, which make it nigh on impossible to care successfully in some cases (consider, for instance, same-sex partners in societies where homosexual relationships are forbidden). Societal oversight of care can thus perpetuate injustice and actively harm those in need of care. It can lead members of some communities to be neglected, exacerbating existing issues and creating new ones. Tronto rightly points out that care work is both undervalued and done primarily by the underprivileged (Tronto Reference Tronto2013). Because they are so undervalued, caregivers are denied social mobility, and their own families and communities are denied care. This is not to say that care's immersion in a complex web of power relations is always to its detriment. We expect nurses and teachers to be held to certain professional standards because we want the vulnerable to be protected to some extent by institutional standards for caregivers’ conduct. Indeed, it seems to be the case that structuring care will often make it easier. In subjecting certain caring practices to sets of norms, we afford caregivers a set of practical and epistemic shortcuts by making particular decisions for them and by pointing them in particular deliberative directions. Professional codes of conduct, for instance, enshrine prohibitions that enable smoother decision-making for those occupying the role of caregiver in morally thorny situations. Thus, when these issues are dealt with well, healthcare professionals do not have to shoulder the burden of moral decisions about confidentiality or patient autonomy—in such cases the imposition of norms by entities external to the caring relation are beneficial. The point here is simply that regarding care as a practice sensitizes us to these sorts of power dynamics.

A related, less explored point here is that definitions of care, good care, and particular subcategories thereof are often subject to power dynamics. As MacIntyre notes, in taking on social dimensions, practices are subjected to communal norms or standards of excellence (MacIntyre Reference MacIntyre2014). Good care is no different. The gold standard for care is, in many practices, determined in large part by a community of practitioners or supervisors, sometimes with little or no consultation with the care-recipients (consider here recent debates over what good care of detainees at the southern US border consists in). The dangers here are both obvious and grave. To act caringly cannot be to act paternalistically in trying to meet needs or without concern for the intended beneficiary's own perspective. Feminist philosophers in general have been adamant that minorities and marginalized groups ought not to have their needs determined for them. Nancy Fraser complains of “struggles over the who and how of need interpretation, struggles to empower women to interpret their own needs and to challenge the anti-participatory, monological practices of the welfare system” (Fraser Reference Fraser1987, p. 118). Although some professionals, such as doctors, might have special insights into what is required for human flourishing, it does not follow that they are omniscient, as though they could perform perfectly caring actions without ever consulting those for whom they care. For though they might have unique knowledge about what some element of our flourishing requires, it is often the care-recipients themselves who know what will best promote their overall flourishing. Even caregivers with special insights must be attentive and willing to listen and learn if they are to best promote the flourishing of their charges.Footnote ⁹ Concepts of practice remind us not only that care's success is determined by external power relations, but also that our notion of successful care can also be determined by those relations, and consequently they remind us that we ought to be cautious about such moral-hierarchical authority.

It is also interesting to note that practices are often thought to involve benefiting one's community in some sense. MacIntyre's understanding of communal benefits is restricted to that community of people who partake in a particular practice. Building on his discussion of internal goods, MacIntyre writes:

Similar examples can be thought of in caring contexts. Florence Nightingale excelled in her field and as a result revolutionized nursing, enriching the medical profession and the lives of those dependent upon it. Nightingale's example will be interesting also to those who think that practices contribute not only to the good of the community of practitioners, but also to the good of the wider community. The practice of care is noteworthy because, unlike some MacIntyrean practices (chess, for example), we depend upon it to meet basic survival needs, and these everyone has. Excellent care, as we see in Nightingale's case, can be a contribution to one's society or to the community of the needy as a whole, and if we treat care as a practice, we introduce this as an additional yardstick for the evaluation of caring actions (though not necessarily as a yardstick for those features of the action for which the agent is responsible, since other social/political impediments may exist).