Separateness of Perspectives, Separateness of Persons, and Duties to the Self

I. SCHOFIELD’S CASE FOR DUTIES TO THE SELF

Schofield begins his case for the existence of duties to the self by sketching out a broadly Darwallian account of morality as “second-personal.”⁴ Stephen Darwall characterizes the second-personal standpoint as “the perspective you and I take up when we make and acknowledge claims on one another’s conduct and will.”⁵ To shed light on this idea, consider Darwall’s recurrent example of a person who steps on another person’s foot. The stepper has a moral duty to remove her foot. One way to make sense of this is third-personally: the stepper would best promote overall happiness by removing her foot. But, thinks Darwall, this would miss the important sense in which the stepper owes it to the other person to remove her foot. The situation is thus better understood second-personally: the stepper is accountable to the victim, her reason to remove her foot is generated by the victim’s legitimate claims or demands, she owes it to the victim to remove her foot, if she fails to remove her foot she wrongs the victim, and so on. This sort of analysis can be generalized to other cases that involve moral duties. So, on the Darwallian view, the distinctive feature of morality is that it involves a “circle of irreducibly second-personal concepts”—concepts such as accountability, legitimate claims and demands, owing, wronging, and so on.⁶

Schofield then argues that a person can relate to herself in this sort of second-personal manner. To do so, he makes the intriguing and provocative suggestion that what allows two distinct persons (such as the foot-stepper and her victim) to relate second-personally is not their being metaphysically distinct persons per se but rather their occupying and deliberating from distinct perspectives with conflicting interests. Of course, distinct persons do in fact always have distinct perspectives, but it is the conflicting perspectives, rather than the metaphysical distinctness of persons, that is philosophically central to the second-personal standpoint—or so Schofield contends. In other words, it is possible to describe a recognizably second-personal interaction without making any reference at all to metaphysically distinct persons.⁷

This means that if one and the same person can occupy and deliberate from distinct perspectives with conflicting interests, she can properly be said to relate to herself second-personally in the way that is distinctive of morality on the Darwallian view. These conflicting perspectives would enable each of us to hold ourselves accountable and make demands on ourselves in the same sort of morally inflected way in which we hold others accountable and make demands on them. There would thus be an important analogy between interpersonal and intrapersonal moral address—between relating to others second-personally and relating to oneself second-personally.

There are two main ways, says Schofield, in which one and the same person can occupy and deliberate from distinct perspectives with conflicting interests. The first and most obvious way is that each of us can occupy and deliberate from distinct temporal perspectives. Consider a young person’s decision about whether or not to smoke. From her perspective as a young person, she experiences smoking as pleasurable. Yet from her perspective as a sufferer of lung disease, she feels very differently about smoking. Here the interests that a person has from distinct temporal perspectives come into conflict in much the same way that the interests of metaphysically distinct persons do. And so it seems perfectly natural that when deliberating about whether or not to smoke, this person might adopt a second-personal standpoint towards herself. That is, qua occupant of the future perspective she might well confront herself qua occupant of the present perspective, holding herself accountable and demanding that she does not ruin her lungs. And, feeling the normative pressure of the demand emanating from this future perspective, she might well conclude that she ought not to smoke. In this way, she would be acting not for reasons of prudence and rational self-interest, nor even for third-personal moral reasons, but rather for distinctively second-personal moral reasons having to do with what she owes to her future self.

The second main way in which one and the same person can occupy and deliberate from distinct perspectives with conflicting interests involves our various practical identities. The interests and ends associated with our various practical identities can come into conflict, just as they do in cases involving metaphysically distinct persons and distinct temporal perspectives. For instance, one’s interest qua parent in spending time with one’s child might conflict with one’s interest qua philosopher in traveling to conferences. This again makes it possible for a person to relate to herself in a recognizably second-personal way: at any given moment, she can issue demands from the perspective of one of her practical identities to the perspective of another of her practical identities. This sort of phenomenology is pervasive. We often talk about feeling “torn” between different aspects of ourselves. “Such talk,” Schofield says, “amounts to an acknowledgement of the possibility of a divided self—a self susceptible to normative pressure applied in opposed directions from competing demands, issued from various perspectives that it occupies.”⁸

Importantly, Schofield’s account of duties to the self solves the paradox mentioned above. That paradox, recall, is that a person who owes a duty to herself would be able to waive that duty, in which case it would lack the binding force characteristic of moral duties. In the case of moral duties arising from the conflicts between distinct temporal perspectives, this objection is defused because any successful waiving of a duty would have to come from the perspective from which that duty was generated. But a person cannot waive a duty generated from a future perspective as she does not yet occupy that perspective. In the case of moral duties arising from the conflicts between the distinct perspectives of our various practical identities, matters are more complicated but the key point is that there are certain practical identities that simply cannot be relinquished or abandoned at will. One way in which Schofield explains this is by drawing on the work of Kantians such as Korsgaard who hold that “our identity as rational beings” is one that “is had by each of us necessarily.”⁹ In both cases—those involving duties across time and those involving duties at a moment—the paradox is solved since duties to the self are binding and unwaivable.

Having defended in this way the idea of moral duties to oneself, Schofield argues that at least some of these duties are enforceable by political institutions. Here he draws on contemporary republican theorists such as Philip Pettit, who hold that state power must be aimed at ensuring that no citizen is dominated by any other.¹⁰ There are different specifications of non-domination, but the core idea is that no one should be subject to the will of another. For Schofield, this non-domination rationale for coercive enforceability applies intrapersonally as well as interpersonally. So just as certain laws are necessary and legitimate in order to prevent one from relating to others in a dominating or unjust way, certain laws are necessary and legitimate in order to prevent one from relating to oneself in a dominating or unjust way. Consider again the choice of a young person to smoke. “From her perspective later in life,” Schofield explains, “such a person seems to be at the mercy of her earlier self—her purposiveness seems to be vulnerable to decisions she makes earlier on—in much the same way that an individual can come to be at the mercy of another person who is in a position to interfere with her body. This explains how an individual person, by virtue of her temporal extension, can relate to herself wrongfully.”¹¹ A paternalistic law banning smoking might thus be justified on the grounds that it prevents this sort of wrongful bodily domination.

While Schofield’s defence of paternalism may differ from other defences of paternalism in some of its practical conclusions about law and policy, the much more important difference is in the nature of the argument used to reach those conclusions. Two distinctive features of this argument are worth noting. The first is that it does not relying on balancing justice against well-being. Many paternalists argue that there are justice-based costs to paternalism but that these costs are outweighed by the benefits of promoting citizens’ well-being. Many paternalists, that is, argue that citizens’ interests take precedence over their independence in these cases. Richard Arneson, for instance, refers to this sort of “conflict within the soul of liberalism” and states that when “paternalism could improve a person’s life dramatically, perhaps along many dimensions of assessment, but at the cost of a small loss of integrity,” it is all-things-considered justified.¹² By contrast, Schofield’s argument is that paternalism is not separate from, and competitive with, justice. Rather, paternalism is constitutive of justice: it is “necessary for putting a person on [just and] rightful terms with her own self.”¹³ Put differently, it is necessary to ensure a fair distribution of the benefits and burdens of “intrapersonal cooperation” between the distinct perspectives within a person’s life.¹⁴ As such, paternalism does not even have pro tanto justice-based costs—paternalistic coercion is no more presumptively unjust than coercion aimed at, say, preventing people from harming others.

The second distinctive feature of Schofield’s argument for paternalism concerns its lack of assumptions about citizens’ rational capacities or dispositions. Many paternalists argue that paternalistic laws and policies are required in light of evidence from the fields of social psychology and behavioural economics that indicates that human beings are systematically and predictably subject to various sorts of cognitive biases and failures of rationality. Sarah Conly, for instance, says in this regard that “the truth is that we don’t reason very well, and in many cases there is no justification for leaving us to struggle with our own inabilities and to suffer the consequences.”¹⁵ This has, in turn, been critiqued by anti-paternalists such as Jonathan Quong who worry that paternalism “demean[s]” or “denigrate[s] citizens’ moral status” insofar as it relies on “negative assumptions about citizens’ abilities to know and rationally pursue their own good.”¹⁶ By contrast, Schofield need not make any such assumptions. This is because non-domination has a requirement of counterfactual robustness: it requires not just that no one actually subjects you to their will but also that no one could subject you to their will. So just as the creation of a society in which all enjoy non-domination requires the state to enact laws that prevent citizens from treating others unjustly even if citizens have no actual desire or tendency to treat others unjustly, it also requires the state to enact laws that prevent citizens from treating themselves unjustly even if citizens have no actual desire or tendency to treat themselves unjustly.

Such, in bare bones, is Schofield’s account of moral and political duties to the self. Before developing an objection to this account, two clarifications are in order. The first is about form versus content. Schofield is at pains to stress that he is not seeking to defend any particular moral or political duties to the self.¹⁷ He does not try to adjudicate between various possible principles for the regulation of one’s relationship with oneself, such as an intrapersonal difference principle (whereby any inequalities between perspectives must be to the greatest benefit of the worst-off perspective), an intrapersonal prioritarian principle (whereby benefitting a perspective matters more the worse-off the perspective is), an intrapersonal sufficientarian principle (whereby each perspective must have enough), and so on.¹⁸ Instead, he is seeking to defend a general framework for thinking about these matters in a fruitful and illuminating way. In this sense, his focus is on the form of intrapersonal morality and justice, rather than its substance and content. In this regard, Schofield explains that working out the substance of intrapersonal morality and justice—working out the substantive moral and political principles that regulate a person’s relationship with herself—is a “prodigious undertaking” and one that is likely to be just as difficult as the task of working out the substance of interpersonal morality and justice.¹⁹ So what still needs to be identified—and what Schofield leaves deliberately open—is a scheme of plausible principles of intrapersonal morality and justice.

Second, it is worth clarifying what is at stake in this debate. After all, even critics of the idea of duties to the self readily accept that we have prudential reasons to promote our own well-being, to develop our own talents, and so on. They just deny that these prudential reasons amount to moral reasons and moral duties. So isn’t this a merely verbal dispute—a dispute about how to use the terms “morality” and “moral duty”? Why does it matter whether we say that treating ourselves in certain ways is required by morality or by prudence, if everyone agrees about how we should treat ourselves? Schofield gives several reasons why it matters whether our treatment of ourselves falls under the rubric of morality as opposed to mere prudence, but the most important of these concerns the distinctive strictness of moral duties. Whereas norms of prudence and rationality “afford a degree of latitude to persons laboring under them,” norms of morality leave little or no discretion.²⁰ Morality “doesn’t simply urge, it commands.”²¹ Whether or not there are moral duties to the self can thus be expected to have far-reaching implications for practical deliberation.

II. INTERPERSONALITY AND ITS MORAL SIGNIFICANCE

Schofield’s argument for moral and political duties to the self relies on some quite heavy machinery, such as Darwall’s second-personal standpoint, Korsgaard’s idea of our inescapable practical identity as rational beings, and republican theories of freedom as non-domination. A perfectly reasonable response here would be to deny one of these theses—that is, to deny that morality is fundamentally second-personal, to deny that we have certain inescapable practical identities, or to deny the coherence or plausibility of republican freedom.²² But, to get to the most original and distinctive issues raised by Schofield’s argument, I shall be maximally concessive: I shall argue that, even if we grant the heavy machinery that Schofield relies on, his argument still fails because it does not take seriously enough the moral significance of interpersonality.

For Schofield, the way in which we relate to others resembles the way in which we relate to ourselves. Just as we issue demands to others, hold others accountable, experience conflicts of interest with others, confer benefits on others, impose harms on others and dominate others, so too each of us occupies multiple perspectives (across time and at a moment) from which we can and do issue demands to ourselves, hold ourselves accountable, experience conflicts of interest within ourselves, confer benefits on ourselves, impose harms on ourselves and dominate ourselves. If the first set of interactions is regulated by moral and political norms, why shouldn’t the second set be regulated by moral and political norms too?

There is an obvious (perhaps too obvious, as I explain below) answer to this question: namely, that interpersonality is itself the relevant difference between the first set of interactions and the second set of interactions. On this view, the reason why interpersonal interactions are regulated by moral and political norms whereas intrapersonal interactions are not is simply that the former involve interactions between metaphysically distinct persons whereas the latter do not—or, at least, something very close to this, such as the fact that interpersonal interactions involve interactions between the distinct wills of metaphysically distinct persons whereas intrapersonal interactions do not.

The point—to be clear—is not conceptual or definitional. It is not that “morality” by definition concerns the interactions between metaphysically distinct persons (or wills) and thus that the idea of intrapersonal morality involves a sort of conceptual confusion. That way of dismissing duties to the self would be deeply unsatisfying. After all, Schofield could easily respond by introducing some novel term (“schmorality”) that has all of the characteristics of morality (such as its strictness, its connection to certain reactive attitudes, and so on) except the characteristic of interpersonality. All of Schofield’s questions and arguments could then be coherently rephrased in terms of schmorality: we could coherently inquire into whether there are schmoral duties to the self, whether some of these schmoral duties are enforceable by political institutions, and so on. My argument against Schofield’s case for duties to the self is thus driven by normative rather than conceptual considerations: it is driven by considerations having to do with the moral significance of the metaphysical distinctness of persons.

In a footnote, Schofield does briefly consider the possibility that the intrapersonal case and the interpersonal case are “disanalogous precisely in that the latter case, but not the former, involves two distinct persons.”²³ But he dismisses this suggestion on the grounds that it “would be question-begging”—hence the worry above that interpersonality might be too obvious a disanalogy here.²⁴ As Schofield explains: “My purpose here is to determine whether a person can have self-directed second-personal reasons of the sort she has with respect to others. To reject my argument on grounds that cases I’m considering involve only one individual is simply to assume a negative answer to the very question I’m attempting to ask.”²⁵

However, this dismissal seems to me too quick. No doubt there are certain crude ways of running an interpersonality-based argument that would indeed beg the question against Schofield. But there are also, I think, more sophisticated ways of running such an argument that are not question-begging in any straightforward sense. In particular, I shall in this and the following sections seek to work up an independently persuasive account of interpersonality and its moral significance—an account that is supported by robust intuitive judgements and thus that is capable of explaining in a non-question-begging way why moral and political norms apply between persons but not within persons.

Let us start with a simple pair of cases:

Alf: Alf commits suicide for no good reason and contrary to the objections and demands that would be issued from the perspective of his future self. Alf’s life is going perfectly well, yet he decides anyway to kill himself.

Betty: Betty kills another person for no good reason and contrary to the objections and demands issued from the perspective of her victim. Betty’s victim’s life is going perfectly well, yet she decides anyway to kill this third party.²⁶

Different people are likely to have different intuitions about whether Alf acts wrongly. Some will deem Alf’s act morally wrong, whereas others will deem it imprudent but not wrong. But surely everyone—even those who do think that Alf acts wrongly—will share the intuition that Betty’s act is more seriously wrong than Alf’s act. Whatever we think about the moral status of Alf’s act, that is, it is difficult to believe that Alf’s act and Betty’s act are on a moral par.

This is not an isolated pair of cases. It is possible to construct many such pairs in which it seems more seriously wrong to impose a harm on someone else than to impose an otherwise similar harm on oneself. In doing so, however, we have to ensure that our pair of cases is “clean.” We have to ensure, that is, that the only relevant difference between the cases is the fact that one case is interpersonal whereas the other is intrapersonal. This is not entirely straightforward because our own consent is normally implicit in our treatment of ourselves, whereas others’ consent is not similarly implicit in our treatment of others—and indeed in Section III.A., I consider the objection that the case of Alf and Betty is not equalized in terms of consent. As Muñoz and Baron-Schmitt put this point, in our treatment of ourselves we are normally “an active participant, not an unwilling victim.”²⁷ For instance, it is implicit in my ruffling of my hair that I consent to being touched in this way, whereas it is not implicit in my ruffling of someone else’s hair that he consents to being touched in this way. In constructing pairs of cases, then, we have to find situations in which a person does not or cannot validly consent to her treatment of herself. In addition to suicide versus murder, such cases might involve harming one’s own long-term health versus harming someone else’s long-term health, getting oneself into debt versus getting someone else into debt, accidentally or negligently giving oneself a painful electric shock versus accidentally or negligently giving someone else a painful electric shock, and so on.

What, then, explains this intuitive difference in wrongness between what Alf does and what Betty does? Given that the two actions seem otherwise equivalent—given, that is, that both involve killing someone for no good reason—a natural explanation for this intuitive difference is the fact of interpersonality itself: inflicting a harm on someone else is more seriously wrong than inflicting a harm of the same magnitude on oneself precisely because it is inflicted on a metaphysically distinct person. On this view, the reason Betty’s act is more seriously wrong than Alf’s act is nothing more mysterious than the fact that Betty acts upon a third party whereas Alf acts upon herself.

Importantly, this disanalogy seems to be morally relevant: the “separateness of persons” is not just an arbitrary difference between the cases (like the fact that Betty’s name is longer than Alf’s) and instead seems on reflection to possess a certain independent plausibility and thus to be of the right kind to explain why the interpersonal case is morally inflected whereas the intrapersonal case is not. It is telling in this regard to observe that when Darwall characterizes the second-personal standpoint, he typically makes explicit reference to the separateness of persons: the second-personal standpoint, he says, is “the perspective you and I take up when we make and acknowledge claims on one another’s conduct and will.”²⁸ Similarly, when Pettit characterizes republican non-domination, he too makes explicit reference to this feature: being dominated consists in “being subject to the potentially capricious will or the potentially idiosyncratic judgement of another.”²⁹ Of course, Schofield would contend that these references to the idea of interpersonality are mere expository artefacts. We can abstract from this idea without losing sight of anything very important. And, admittedly, Darwall and Pettit do elsewhere characterize their positions without explicit reference to the idea of interpersonality and thus in ways that leave open the possibility of moral and political self-relation. Pettit, for instance, elsewhere says that “to enjoy non-domination is to be in a position where no one [oneself included?—CT] has that power of arbitrary interference over me.”³⁰ Still, the separateness of persons does seem to mark a deep and important moral fact, and it is at least suggestive that the very theorists on whom Schofield relies in building his case for duties to self often appear to regard this fact as morally relevant.

Many others also make use of the idea of the separateness of persons. Robert Nozick, for instance, says that individual moral rights “reflect the fact that no moral balancing act can take place among us.”³¹ He continues: “there is no moral outweighing of one of our lives by others so as to lead to a greater overall social good. There is no justified sacrifice of some of us for others. This root idea, namely, that there are different individuals with separate lives and so no one may be sacrificed for others, underlies the existence of moral side constraints.”³² Similarly, Rawls famously criticizes classical utilitarianism on the grounds that it “adopt[s] for society as a whole the principle of rational choice for one man” and so “does not take seriously the distinction between persons.”³³ As these statements indicate, the separateness of persons is typically used to show that the interpersonal cannot be modelled on the intrapersonal (as utilitarians endeavour to do). But it can equally well be used to show that the intrapersonal cannot be modelled on the interpersonal (as Schofield endeavours to do). Attempts to analogize the interpersonal and the intrapersonal—whether in the service of utilitarianism or in the service of duties to the self—fail to take seriously the normative significance of the separateness of persons.

The separateness of persons is particularly important from the point of view of liberal political philosophy.³⁴ Indeed, for many liberal theorists interpersonality is part of the nature and point of politics. Liberal politics (even more than liberal morality) is fundamentally interpersonal: it is about creating a framework in which citizens are free to lead their own lives in their own way so long as they do not harm others,³⁵ or in which citizens enjoy independence from the wills and choices of others,³⁶ or in which citizens despite their conflicting ends can cooperate with each other on fair terms as free and equal persons,³⁷ or something in this vicinity. Despite their differences, these liberal theories agree that politics is concerned with specifying the normative relations between persons—hence the popularity amongst liberal theorists of the metaphor of a social contract as opposed to a Ulysses contract. Liberal theory thus leaves little if any room for relating to oneself politically. Arthur Ripstein, for instance, says in this vein that “Kantian independence is not a feature of the individual person considered in isolation, but of relations between persons … [It] can only be compromised by the deeds of others.”³⁸ Schofield’s denial of the relevance of interpersonality—and his corresponding receptiveness to enforceable duties to oneself to prevent certain forms of political self-relation such as intrapersonal harm, intrapersonal domination, intrapersonal inequality, intrapersonal injustice, and so on—thus constitutes a striking departure from the dominant liberal conception of the nature and point of politics.

In short, the fact of interpersonality offers a straightforward and independently plausible explanation of the intuitive moral difference between the wrongness of Alf’s action and Betty’s action (as well as other similar pairs of cases)—an explanation that is especially congenial to the liberal tradition. Insofar as the apparent moral difference between this pair of cases (i) cannot be better explained in some other way and (ii) is a genuine moral difference, the fact that interpersonality accounts for this difference constitutes evidence in favour of the moral relevance of interpersonality and thus against the attempt to analogize interpersonal address and intrapersonal address. In the following sections, I will defend in turn each of these suppositions against objections.

III. EXPLAINING THE DIFFERENCE WITHOUT APPEALING TO INTERPERSONALITY

I have suggested that the intuitive moral difference between what Alf does and what Betty does is explained by the fact of interpersonality itself. But that explanation, of course, is unavailable to Schofield, who does not attribute any intrinsic moral significance to interpersonality. So unless he wants to deny that Betty’s act is more seriously wrong than Alf’s act (an option discussed in the next section), he will need to provide some alternative explanation of this difference—one that does not make essential reference to the fact that Alf harms himself whereas Betty harms someone else. There are at least three ways in which Schofield might explain this difference without appealing to the fact of interpersonality. I call these the “consent” strategy, the “different-principles” strategy and the “different-strictness” strategy. None, I shall argue, is successful.

IV. DENYING THE DIFFERENCE

Alternatively, Schofield might bite the bullet and deny that there is in fact any genuine moral difference between what Alf does and what Betty does. On this view, Alf’s act and Betty’s act are on a moral par insofar as both acts inflict a serious harm an unwilling victim.

However, this bullet-biting strategy comes at a very significant cost. The intuition that gratuitous murder is more seriously wrong than gratuitous suicide is extremely robust. Indeed, the very fact that our language as well as a wide range of other languages has a special word for suicide—as opposed to assimilating suicide into the category of murder—is an indication of the depth of this distinction within common-sense morality.⁴⁸ And while it is true that some people do use the word “self-murder,” this usage is very rare: for most people, suicide is not murder. Moreover, if Schofield were to say that there is no moral difference between gratuitous murder and gratuitous suicide, he would probably also need to say—even more counterintuitively—that there is no political difference here, since duties not to kill or cause serious bodily harm are paradigmatic examples of politically enforceable duties. Yet the vast difference in punishment for the crime of attempted murder and the crime of attempted suicide—where indeed the latter is considered a criminal offence at all—across a wide range of jurisdictions suggests that murder and suicide are not normally thought to be on a political par in this way. None of this, of course, is intended to trivialize suicide. The point is just that however wrong suicide is—and it may well be very seriously wrong—murder is widely thought to be even more seriously wrong.

It is not enough, then, to simply assert that what Alf does and what Betty does are on a moral par. For such a strategy to be persuasive, one would need to say much more in support of this assertion. In doing so, Schofield might again stress that what ultimately matters is not the separateness of persons but the separateness of perspectives. And since both Alf and Betty kill someone for no good reason and against the objections issued from the relevantly affected perspective, aren’t their actions equally wrong? How could it make a deep moral difference—Schofield might ask—whether a conflict between distinct perspectives occurs within a person or between persons? Admittedly, we are used to thinking in terms of distinct persons rather than distinct perspectives. But why, on more abstract reflection, should that matter? In particular, Schofield might stress that “a person can confront her own will as foreign, just as she might confront the will of another, by virtue of the fact that it can knowingly undermine itself in much the way that distinct wills can.”⁴⁹ Given this possibility of inner conflict and alienation from one’s own will, can the bare fact of interpersonality bear the moral weight placed upon it? Why should I care whether I am subject to someone else’s will or subject to my own will, if in both cases I am subject to a will that I strongly oppose? Shouldn’t comparable harms be equally wrong, regardless of whether they flow from the will of someone who is literally other or from the will of someone who is metaphorically other?

I agree that a person can confront her own will as foreign—or at least that a person can confront some aspect or portion of her will as foreign. But, still, there remains a phenomenologically important difference between the sort of foreignness that a person can feel in relation to her own will and the sort of foreignness that she can feel in relation to someone else’s will. Even when a person confronts her own will as foreign or alien in certain respects, she confronts it as less foreign than the will of someone else. She still identifies with her will; she is intimately associated with it; she recognizes it as hers in some normatively weighty sense.⁵⁰

Of course, Schofield might say that this phenomenologically important difference between being harmed by my will and being harmed by someone else’s will does not correspond to any morally important difference. But this cannot be right. When a person ignores her future well-being, or undermines herself, or wastes her resources, it is appropriate for her to come to regret this, or feel guilty about this, or take responsibility for this.⁵¹ All of these attitudes and emotions suggest a sense of identification with one’s past will that coexists alongside a sense of alienation from that will. They point towards the existence and significance of an enduring self—a first-personal identity to which a person’s various conflicting perspectives and standpoints belong. By contrast, it does not make sense for a person to regret, or feel guilty about, or take responsibility for the harms that other persons impose on her. These attitudes and emotions are incoherent, or at least inappropriate, in relation to the decisions of third parties. These sorts of differences in the appropriate attitudinal and emotional responses to our own choices versus others’ choices strongly suggest that the fact of interpersonality—the fact that I am harmed by my own will as opposed to someone else’s will—has moral significance and not merely phenomenological significance.

In short, then, while a person can confront (aspects or portions of) her own will as foreign, she still confronts her own will as significantly less foreign than the will of someone else. Even when we oppose our own past and present choices, we appropriately identify with those choices in a way that we do not identify with others’ choices. So appealing to the possibility of inner conflict and alienation from one’s own will does not succeed in undermining the moral relevance of interpersonality or in showing that Alf’s act and Betty’s act are on a moral par.

V. SHEDDING LIGHT ON THE SEPARATENESS OF PERSONS

I have argued that Schofield’s argument for the existence of moral and political duties to the self, despite its ingenuity and sophistication, does not adequately account for the most central disanalogy between interpersonal interactions and intrapersonal interactions—namely, the fact of interpersonality itself. He suggests that it would be question-begging to cite this disanalogy—for this would be to deny precisely what he asserts. But I have sought to develop an independently compelling (and thus non-question-begging) account of interpersonality and its moral significance—an account informed by self-standing intuitive judgements about pairs of cases. However, beyond posing a challenge to Schofield—who must now either explain away the intuitive difference between the wrongness of gratuitous suicide and the wrongness of gratuitous murder or else explain this intuitive difference without relying on the fact of interpersonality itself—this argument provides several independently interesting insights into the idea of the separateness of persons. Of these, I shall mention two: the first concerning the explanatory role of the separateness of persons, and the second concerning the first-order implications of this idea.

First, then, the argument in the preceding sections helps us to sharpen our understanding of the explanatory role of the idea of the separateness of persons. Many moral and political philosophers appeal to the idea of the separateness of persons in order to support or criticize a given view. But these philosophers do not typically explain how this idea fits into the normative landscape. At what level of fundamentality is the idea of the separateness of persons most plausibly thought to operate? Is this idea explanatorily fundamental, or at least close to being explanatorily fundamental? Or is it instead downstream, in some sense, from various even deeper values, conceptions and commitments? And, if the latter, what are these more basic values, conceptions and commitments, and how do they relate to the idea of the separateness of persons?

Schofield takes the views of various theorists who appear to attribute importance to the idea of the separateness of persons—theorists such as Darwall and Pettit—and seeks to show that, when one follows through on these views, what turns out to be of fundamental importance is not the separateness of persons per se but instead the separateness of perspectives. This shift from the separateness of persons to the separateness of perspectives, I have contended, should be resisted—for, otherwise, we face the absurdity of treating Alf’s action and Betty’s action as morally equivalent. And the only way of resisting this shift, it turns out, is to treat the separateness of persons as even more fundamental than we might previously have imagined. In this way, reflecting on the idea of the separateness of persons in the light of Schofield’s project teaches us something independently interesting and important about this idea’s explanatory role and status within moral theory: it teaches us that the separateness of persons is something like a brute fact about morality. Ultimately, “the self is simply special.”⁵²

This is not to say that nothing can be said in support of the moral relevance of the separateness of persons. After all, unless we regard the separateness of persons as morally relevant, we cannot explain our intuitive judgements about the difference in wrongness between gratuitous suicide and gratuitous murder (as well as the difference in wrongness between other similar pairs such harming one’s own long-term health and harming someone else’s long-term health)—or so I argued in Section III. Within the methodology of reflective equilibrium, these sorts of considerations of coherence with intuitive judgements about cases do in themselves lend support to the moral relevance of the separateness of persons. The point is just that the moral relevance of the separateness of persons cannot be defended by reference to any more general or immediately intuitive idea from which it can be derived or under which it can be subsumed (such as the separateness of perspectives). It is a fact of irreducible moral significance—an axiom rather than a theorem. Given the obscurity that sometimes surrounds the idea of the separateness of persons, this sort of clarity about its explanatory role and status within moral theory is welcome.

Second, the argument in the previous sections sheds light on the first-order implications of the idea of the separateness of persons. In particular, the argument suggests that a morality that takes seriously the separateness of persons is fundamentally patient-relative. Moral philosophers have long been familiar with the idea of agent-relativity—the idea, that is, that the moral status of an action can depend on whether I do the action or whether someone else does the action.⁵³ This idea figures prominently in debates over consequentialism and deontology. As Nagel puts it, “deontological reasons have their full force against your doing something—not just against its happening.”⁵⁴ But the argument in the preceding sections points towards the much more unusual and unfamiliar idea of patient-relativity—the idea, that is, that the moral status of an action can depend on whether I do the action to myself or whether I do the action to someone else.

In a sense, of course, the idea of patient-relativity is not so unusual or unfamiliar: it is another name for the common thought shared by many moral and political philosophers that there are no (or almost no) duties to self, or at least that there is a significant difference between what we owe to others and what we owe to ourselves (that is, the “self-other asymmetry”). But this name helpfully draws attention to a deeply puzzling aspect of that common thought. For it is strange to think that I can free myself from general moral constraints against harming just because I am me. Isn’t that a kind of special pleading? How can morality allow me to make an exception for myself in this way? Isn’t this just as strange as thinking that I can make an exception for you because you are you? Surely morality shouldn’t allow differential treatment based on who the patient is?⁵⁵ If the preceding argument is sound, however, morality is indeed patient-relative in this way. Morality allows differential treatment based on who the patient is in the case of oneself and oneself alone: it allows me to make an exception for myself (from, say, general constraints against harming) because I am me, but does not allow me to make an exception for you because you are you, and there is no deeper explanation of why this is so, beyond the fact that I am me and you are you.

Competing Interests

The author declares that he has no competing interests.