1.A. Wrongs Without Rights

Cornell sets up a dilemma for the rights-theorist—who wants to maintain that a person is wronged if and only if their right is violated—by constructing a series of cases where an agent A wrongs C by either (i) violating B’s right, or (ii) violating a general standard of conduct, or (iii) acting badly or imprudently. In these kinds of cases, the rights theorist must either (a) say that C had a right, the violation of which explains why A wronged them; or (b) deny that C was wronged (because no right of theirs was violated). But Cornell argues that neither response is plausible: To say C had a right in these cases implausibly stretches the concept of a right, and to deny that C was wronged is just to deny manifest reality. The inescapable conclusion is that A wrongs C. And since A can wrong C without violating C’s rights, then there is a sense in which wrongs and rights come apart.

Cornell offers three kinds of cases. The first kind involves a third party who is wronged by the violation of a second party’s rights: A violates B’s right and thereby wrongs C (27).

Cornell’s second species of case drops rights violations altogether. These are cases where A violates some “general public norm”—including “the demands of the law, morality, or simply ‘human need’”—and thereby wrongs C (51). Cornell’s view is that, “as a general matter, we may be wronged by wrongful conduct that affects us” (49).

Cornell’s third kind of case ‘speculatively’ expands the claim one step further and drops the requirement that A acts wrongly at all. Sometimes A acts imprudently and thereby wrongs C. “Because others have a stake in our lives and our actions,” Cornell contends, “our imprudence and poor choices can wrong them” (52).

There are two general strategies for avoiding the conclusion that these cases are instances where C is wronged without a violation of their rights. The first is to insist that any time we are inclined to agree that C is wronged, it is because they had a right against the wrongful conduct. The second is to deny that C was wronged at all, because they did not have a right against the conduct that caused them harm. Cornell argues that neither response is plausible.

This argument largely hangs on Cornell’s stipulated interpretations of the nature of rights and wrongs. Horn one of the dilemma rests on Cornell’s view—largely assumed here, although defended in previous work—that rights are necessarily ex ante reason giving.Footnote ⁸ A right gives the rightsholder control over another’s action. Typically (though not necessarily) a rightsholder can waive or alienate a right, and can enforce or make demands grounded in the right. And rights usually impose deontic requirements on others. For that reason, we should not be promiscuous in assigning rights. Cornell has previously suggested that the concept of a right can “be softened to capture simply a standing to be wronged, at which point it has lost its solid form” and become “buttery.”Footnote ⁹ But when a right is understood simply as a way to say that a wrong was done, the concept loses its explanatory force. Cornell argues that we can attribute a right to C in the above cases only if we attribute to them a “buttery” right, but not a “solid” right.

The sting of the second horn of the dilemma depends on Cornell’s account of wrongs. In lieu of an analysis of wronging, Cornell gives us a heuristic set of marks.Footnote ¹⁰ To determine whether someone is wronged, we need to attend to the kinds of responses that seem appropriate. As Cornell sees it, we are in the vicinity of a wrong when some of the following responses are fitting: (i) the second-person reactive attitude of resentment; (ii) the first-person reactive attitude of guilt; (iii) complaint; (iv) apology; (v) some form of remedial action like compensation or repair; (vi) forgiveness (12). Cornell reports the intuition that many of these responses would be appropriate in the above cases, and that any attempt to deny that C is wronged is hard-hearted, reflecting a “theory-driven refusal to engage with actual life” (43).

1.B. Wrongs Without Reasons

Cornell’s second argument that wrongs and rights come apart is that wrongs, unlike rights, do not always give ex ante reasons for action (as opposed to ex post reasons for repair). This argument depends on the following claim:

Or as he sometimes puts the point more simply, “potential wrongings do not provide any reasons” (69). Cornell supports this claim again by appealing to intuitions about examples. We are invited to share the intuition that in these cases, the fact that φ-ing will wrong but ρ-ing will not wrong gives no reason to ρ rather than φ.

The first cluster of cases involves promises that do not create obligations but the violation of which wrongs the promisee. In other words, A wrongs C by φ-ing, but has no obligation not to φ (70, 74, 76, 78).

In the first cases, A promises something beyond their control: A state of affairs will obtain, or they will feel or believe something specific. But A cannot have an obligation to bring about something beyond their control. A nevertheless wrongs C when what was promised does not occur. Sexual Promise has a different structure, but the same basic idea: A’s promise does not create an obligation because of a fact about sex: Whether a person should have sex is, Cornell asserts, “entirely responsive to whether she wants to have sex” (79). Nevertheless, Cornell invites the reader to share his intuition that the telltale signs of wronging—resentment, complaint, apology—will be present if A breaks the non-binding promise.

Next, Cornell presents a class of cases where an agent faces a practical conflict, with equally weighted reasons on the side of two actions, and where the fact that one action will do wrong does not intuitively generate a reason that breaks the tie (81).

Cornell has the intuition that it is permissible for A to save either B or C and that A does not have a special reason to save B, who is in danger because of A’s negligence, over C, who is in danger because of a freak accident. The fact that A wrongs B but not C provides no reason for A to save B instead of C.

Wrongs Give No Reasons is, in one respect, highly counterintuitive. As I will discuss in Part 3, it seems obvious that one way of fully respecting our deep interconnectedness to others is to treat the fact that we might wrong them as a reason in practical deliberation. To deny this is to adopt a heavy explanatory burden. In another sense, the conclusion that Cornell supports with this claim seems to be just that wrongs and rights possess different properties and so are not equivalent. If the claim that “wrongs and rights come apart” is just the claim that “wrongs and rights are not the same,” then it would not be interesting.

But Cornell’s effort to pull apart our ex ante obligations from our ex post remedial relations is interesting. At the heart of the book is a novel account of moral deliberation and moral community rooted in just this difference between the kinds of rights-grounded reasons for action we have ex ante and the wrongs-grounded reasons for repair we have ex post. As noted in the Introduction, these accounts are meant to displace the Obligation-Accountability Nexus.

1.C. Accountability and Obligations Come Apart

If the claim that wrongs and rights come apart is philosophically significant, it is because it has some significance for our understanding of moral obligation and accountability. In what I take to be one of the most important passages in the book, Cornell indicates why he thinks it is important to separate wrongs from rights:

In this passage, Cornell denies the OAN when he says that our other-directed reasons for action are “neither because of nor mutually constituting with accountability relations.” The problem with the OAN, for Cornell, is that it is frequently appropriate to hold A accountable for failing to φ without excuse even if they had no obligation to φ.Footnote ¹¹ Obligations are determined by rights, while accountability relations are determined by wrongs. Since rights and wrongs come apart, obligations and accountability also come apart.

Rejecting the OAN is significant. It, or something close to it, is affirmed by a range of moral and legal philosophers who are engaged in the same basic project as Cornell: attempting to explain how our social connections to one another are shaped by rights and wrongs—how, in other words, rights and wrongs contribute to the formation and maintenance of a moral community. According to the OAN, what it is to have a moral obligation is that it is appropriate for others to engage in social practices of holding one another accountable, and that each of us is, in some sense, vulnerable to those accountability practices. Jay Wallace puts the idea as follows:

As we have seen, Cornell rejects this picture as an inadequate way of capturing the full scope of our interconnectedness. Obligations do not form the basis of our practices of interpersonal accountability. To think they do is radically hard-hearted, ignoring the many ways that our duties to repair and redress wrongs far outrun our obligations to treat or avoid treating one another in specified ways. If others appear in our moral lives only when we deliberate about what to do, we are missing half the picture.

1.D. The Turn Against Palsgraf

Over the last decade, Cornell has been out front of a growing number of theorists who are broadly sympathetic to non-consequentialist and relational theories of morality and private law, but who call into question various aspects of either the Palsgraf Principle or the OAN.Footnote ¹³ I briefly mention a few entries in this growing literature, before suggesting what is distinctive about Cornell’s approach.Footnote ¹⁴

Brendan de Kenessey has argued that our talk about “directed obligations” is systematically ambiguous between two importantly distinct notions: constraint relations and accountability relations.Footnote ¹⁵ These map pretty well onto what Cornell means by rights and wrongs, respectively. De Kenessey similarly argues that constraint relations and accountability relations come apart: Sometimes A has an obligation to C to φ (grounded in C’s right), but C cannot hold A accountable because A had decisive reason not to φ; and sometimes C can hold A accountable for failing to φ, not because C had a right that A φ but because A’s failure to φ set back C’s welfare interests in an unjustifiable way.

Like Cornell, Jed Lewinsohn aims to “prize apart the determinants of permissible action, on the one hand, and proper aggrievement, on the other.”Footnote ¹⁶ Unlike Cornell, however, Lewinsohn is attracted to a restrictive interpretation of who counts as ‘properly aggrieved’. He is open to some version of a Palsgraf Principle, but denies that there is a coherent individualistic conception of rights that can accurately carve out the properly aggrieved.Footnote ¹⁷ In his view, who counts as properly aggrieved should be determined by “a social or legal policy that is itself justified, i.e., that serves legitimate social goals (e.g., deterring harmful conduct) and fairly distributes benefits and burdens,” including the burden of being held accountable by a properly aggrieved person.Footnote ¹⁸

Finally, Adrienne Martin has defended a related view according to which we have directed obligations that do not correspond to others’ claim-rights, often arising out of what she calls “personal bonds.”Footnote ¹⁹ Yet for Martin, when we violate these directed obligations, it is more appropriate to say that we let down the person to whom it was owed, rather than that we wrong them.Footnote ²⁰ (I will suggest later that I think Martin’s locution is helpful.)

Several features distinguish Cornell’s intervention. First, Cornell has a very expansive understanding of who is properly aggrieved by A’s actions. For Cornell, it is sufficient for C to be properly aggrieved that A acted badly or imprudently and that C had a ‘stake’ in A’s action. Second, as we will see, the Wrongs Give No Reasons thesis plays an important role in Cornell’s argument against the OAN. Third, Cornell offers a distinctive account of moral deliberation and of moral community: On the one hand, we saddle ourselves with a constrained picture of moral community if we focus too much on others’ rights; on the other, it is a mistake to focus in deliberation on the wrongs we might do to others. In what follows, I elaborate and offer some criticisms of each of these elements of Cornell’s view.

2.A. Motivating the Juridical View

Rights-based theories, Cornell tells us, “mistakenly come to believe that the only way that we are tethered to each other is by having rights” (210). In such views, as Cornell would have it, if a person is not wronged in the sense of having their right violated, then that person is no different from a mere bystander: Mrs. Palsgraf is normatively indistinguishable from a journalist at the other end of the train platform who watched it all happen. But such a view overlooks the banal fact that “as important as rights are, they do not exhaust our interpersonal moral relations” (182). Even if the railroad did not violate Mrs. Palsgraf’s rights, it is still specially connected to her in virtue of having wronged her.

I do not think most rights-based theorists would accept this characterization of the position. Rights theorists regularly distinguish between a ‘core’ part of morality and the rest of ethical life. There are many ways to characterize this distinction by singling out a special part of morality. For example, as the part of ethical life that deals with what we owe to each other;Footnote ²¹ or the part that deals with Recht as opposed to Moral or duty as opposed to virtue;Footnote ²² or as ‘deontic’ morality and obligations;Footnote ²³ or as the “morality system.”Footnote ²⁴ Cornell himself acknowledges that there is a “special subset of our duties, the subset concerned with respecting the normative space of others” (213 [footnote omitted]).Footnote ²⁵ For reasons that will become clear, I refer to the singled out part of morality as the ‘juridical’ part. There are many different reasons for singling out a juridical part of morality (I will give what I take to be the most compelling reason below), but I am not aware of any theorist for whom the point of emphasizing the juridical realm is to deny the significance of the rest of normative life. For example, in a passage that Cornell quotes, Arthur Ripstein claims that an example does not involve a wronging “in the sense that is of interest to us here” (48 [emphasis added]).Footnote ²⁶ He thus makes clear that he is not saying there is no sense in which there is a wronging.

What is the point of making this kind of distinction? The purpose is not, as Cornell suggests, to attempt to “cordon[] off the personal from the moral” (56 [footnote omitted]). The point is that there are important differences in kind between the ways we affect one another.

Sometimes we disappoint, frustrate, let down, or confuse one another. These may be morally significant and may call for morally significant responses: a sense of betrayal, disappointment, regret, anger, confusion, pity, self-doubt, resentment, and so on. But sometimes we affect others in a way that implicates their freedom or their status as an equal in the community. And this may call for a distinctive kind of response that in turn implicates the freedom or moral status of the alleged wrongdoer.

When A is liable to C for φ-ing, C can hold A accountable for failing to φ without excuse.

What does it mean for C to hold A accountable? As we will see in the next Section, accountability comes in many varieties, from silently blaming, to heated demands, to civil recourse in court. For now, I will assume that at least some ways that C can hold A accountable threaten A’s freedom (in some sense), even when it is conceded that A has acted badly or wrongly. The most obvious version of this is if C holds A accountable via a coercive legal mechanism like tort remedies. But ‘softer’ forms of accountability can be freedom-threatening too.

Since it can threaten other people’s freedom, the power to hold someone accountable is therefore a highly consequential power. This gives us reason to limit the class of persons who qualify as properly aggrieved, i.e., to whom A can be liable for their actions.Footnote ²⁷ A natural way to restrict the class of properly aggrieved persons is to articulate a juridical domain in which those who are wronged and authorized to hold others accountable in a freedom-threatening way are those whose rights were violated. In other words, for A to be exposed to freedom-threatening liability from C, it is necessary that A violated C’s rights. And it is a short walk from here to the Palsgraf Principle.

2.B. Freedom, Accountability, and the Distinctiveness of Right

Cornell denies that being answerable or accountable to someone for wronging them “entails a lack of freedom” (61). But there are different ways to talk about freedom and about accountability and it is important to clarify which we have in mind. When Cornell insists that “[f]reedom, whatever it means, isn’t about being able to act wrongly without facing others to answer for it,” I think he papers over some important distinctions on which the rights-based approach depends (62). I will sketch a way of making these distinctions rooted in Kant’s concept of a juridical domain of right [Recht]. This is not the only way to do it, but Kant’s approach remains influential and will help fix ideas.

Kant distinguished three kinds of freedom. The first is metaphysical freedom, or the freedom to act at all. This is what is at issue in debates about free will, metaphysical determinism, and the connection between freedom and moral responsibility.Footnote ²⁸ The second is practical freedom, or what Kant called “autonomy.”Footnote ²⁹ This is the capacity to act well, in accordance with right reason. Kant famously worried that we might possess metaphysical freedom but lack practical freedom if, for instance, moral principles cannot motivate action on their own. The third is what Kant called “independence.”Footnote ³⁰ This is the freedom to act without subjection to the coercive power of others.

The concept of accountability is, in turn, sensitive to these distinctions between kinds of freedom.Footnote ³¹ A first kind of accountability has to do with whether it makes sense to attribute actions to a creature lacking metaphysical freedom.Footnote ³² A second way to speak of accountability connects it to morality in the sense at issue when we talk about autonomy and doing what is right. Here we speak about accountability to one’s reasons for actions and accountability to one’s own conscience. We also speak about a kind of accountability to other persons, though not necessarily in a way that implicates ‘directed obligations’. Finally, a third kind of accountability corresponds to the third kind of freedom—independence—and a wholly interpersonal, juridical space of Recht. This kind of freedom is not about our freedom to do well, but rather about our freedom from subjection to others’ power. The corresponding kind of accountability is a vulnerability to enforcement that prima facie subjects us to others’ powers—whether through laws or through ‘softer’ forms of social coercion.

Rights-based theorists are not always careful to distinguish these different senses of freedom and accountability. But at least a plausible version of the rights-based theory is interested in the third kind of freedom and the third kind of accountability. The third kind of freedom as independence might mean being able to act wrongly (in the sense of violating moral duties) without being answerable for it in the third sense of coercive enforcement. But that would not imply anything at all about whether the (moral) wrongdoer is answerable in the first or second sense.

To decide whether A has wronged C, the rights theorist would ask whether C had a right and whether A’s action breached the boundaries of that right. The answer might be “no” even where A’s action violated some norm, causes C harm, and warrants an indignant response from C. Cornell would instead answer whether A has wronged C only by asking whether it would seem appropriate for C to engage in some of the responses characteristically associated with wrongs: resenting, complaining, demanding apologies, retaliating, etc. Cornell recognizes that these responses come in many shades. He distinguishes, for instance, between C’s merely grumbling or feeling resentment, on the one hand, and C’s invoking the coercive powers of the state, on the other. He says the first kind of response does not implicate A’s freedom, but the second might (61). But Cornell insists we focus on the first, non-coercive case, because if we take the class of more ‘serious’ wrongdoing that calls for coercive enforcement as our paradigm, we risk “detaching wrongs from moral dialogue and personal reactive attitudes,” and therefore risk “losing any foundation in ordinary interpersonal life” (62). The idea seems to be that legal remedies tend to be stilted and artificial, while reactive attitudes are closer to the flesh and blood of moral life.

At the same time, however, Cornell concedes that sometimes the expression of reactive attitudes “can be a serious and primal sanction” (62 [footnote omitted]). From this he concludes that it would be “odd to treat the availability of legal remedies as encroaching on our freedom but not to regard reactive attitudes similarly” (62). Cornell evidently means this as a tollens: Both legal remedies and reactive attitudes can involve a ‘serious and primal sanction’. So, either both coercive legal remedies and reactive attitudes encroach on our freedom, or neither do. Since (by assumption) reactive attitudes do not encroach on freedom, neither do legal remedies. This is a counterintuitive conclusion. Cornell tries to make it palatable by pointing out that as long as you act within your rights, exposure to coercive legal enforcement subjects you to “no additional constraints on what you may do” and for that reason does not limit your freedom (62 [emphasis in original, footnote omitted]). But I am not sure Cornell is entitled to this move. He elsewhere maintains that you can wrong others even when you act within your rights, but if wronging someone exposes you to remedial action, then surely that remedial response adds a constraint on what you do by replacing a non-penalized with a penalized option.Footnote ³³ Cornell concedes that there is a “practical sense” in which exposure to coercive enforcement constrains freedom: “One frolics less freely in a space bordered by landmines than in one bordered only by fences” (63). In my view, he undersells how significant this fact is. Even if proliferating those with standing to hold one accountable in a freedom-threatening way does not change one’s first-order obligations, it multiplies the occasions for negotiating the scope of obligations and remedies, i.e., possible occasions of subordination or unfreedom.

More importantly, one person’s tollens is another person’s ponens. The more straightforward view is that both outright coercion and the expression of reactive attitudes at least sometimes raise issues of freedom. We might think that it is precisely because the expression of reactive attitudes can sometimes have a real, practical, censorious, and sanctioning function that they sometimes do encroach on freedom, analogously to the way legal sanctions encroach on freedom. That is not to say legal and non-legal sanctions are identical; they pose somewhat different threats. Legal sanctions tend to be more coercive, but not always: Sometimes we sanction with reactive attitudes in a more ‘serious’ way for more ‘serious’ offenses than the coercive enforcement of the law. Compare the legal enforcement of a $500 contract with the kind of blame that ruins a marriage or excommunicates one from a circle of friends. Moreover, informal non-legal sanctions may be more difficult to appeal and (as we will see below) in some ways more difficult to reconcile with independence.

When a rights theorist denies that somebody was wronged because their right was not violated, I propose this means that the actor is not answerable in the third sense and that trying to make them answerable in that sense would implicate their freedom as independence. Because it implies nothing about the first or second kinds of answerability, there is nothing heart-pinching or inhuman about this kind of view. Cornell’s assertion that Cardozo or anyone who agrees with his rationale in Palsgraf “has mistakenly come to believe that the only way that we are tethered to each other is by having rights,” is therefore unjustified (210). It makes all the difference that in Palsgraf, Cardozo was asked about the railroad’s liability in tort to coercive legal sanctions. He was not asked whether it would be appropriate for Mrs. Palsgraf to resent the railroad’s executives or for them to send Mrs. Palsgraf some flowers to her hospital room with a sincere apology.Footnote ³⁴ Indeed, the rules of evidence in most Canadian provinces and American states already acknowledge this difference: Apologies are often inadmissible as evidence in tort suits.Footnote ³⁵ This allows people to take a kind of interpersonal responsibility without implying anything about freedom-threatening liability.

2.C. The Juridical Case Against the OAN

On the face of it, there is an affinity between the juridical approach just sketched and a principle like the OAN. The OAN says that what it is to have an obligation to φ is to be appropriately held accountable for failing to φ without excuse. Constraint and accountability are run together in a way that serves to limit the class of the properly aggrieved and thereby reduce our exposure to freedom-threatening accountability. Indeed, it is tempting to justify a juridical approach and the Palsgraf Principle by appealing to the OAN or something like it.Footnote ³⁶

But not only is the juridical view not committed to the OAN, properly understood it provides a powerful argument against the OAN as a general principle. The juridical view carves off a special part of deontic morality from the rest of ethical life. This allows us to say that in addition to the indefinite ways we are intimately wrapped up in one another’s actions, there is a special set of cases that involve rights, wrongs, and a special way of holding one another accountable. That special way of holding one another accountable implicates our freedom, understood along the lines of what Kant called ‘independence’. This kind of freedom requires not being subject to the uncontrolled power of another person to interfere with your own rights. Some forms of holding one another accountable—including but not limited to coercive enforcement—threaten this kind of freedom.

This dynamic raises distinctive moral and political questions. Specifically, this kind of freedom-implicating way of holding one another accountable implicates questions of authority. To permissibly hold one another accountable in a way that does not violate each other’s independence, we need authority to do so. But we do not come by this authority for free. As a result, sometimes A has an obligation to do something and when they fail to do it (without excuse), nobody has the authority to hold them accountable for it. This is conceptually incoherent according to the OAN. But it is a familiar feature of moral life. The simplest case is when the content of the moral obligation or the scope of the appropriate response is the object of reasonable disagreement. Where there is reasonable disagreement, C can insist on a freedom-implicating remedy only by holding themselves out as a superior moral authority. But doing so would not be compatible with the alleged wrongdoer’s independence.Footnote ³⁷

Accountability practices require legitimating conditions. But whether those legitimating conditions obtain is contingent. In my view—which I can only state here—the legitimating conditions are best understood as social practices that provide a publicly-acknowledged way to determine both the applicable moral standard, and the norms and procedures of adjudication and for determining an appropriate remedy.Footnote ³⁸ It might be tempting to say that what authorizes C to hold A accountable is just the fact that A wronged C by violating their rights. Or that all persons have an intrinsic authority to hold one another accountable. But that is too quick. The constitutive aim of a juridical system—at least of the kind I am exploring—is maintaining the mutual independence of the parties.Footnote ³⁹ And independence precludes C’s imposition of a remedy on A in the face of reasonable disagreement, even if C is correct and knows it. When C can draw on a shared background of social practices to hold A accountable, they can do so legitimately and with authority. But whether A has a moral obligation is a completely different matter, and it is implausible to suppose that that question turns on the existence of contingent social practices.

Because A can have a moral obligation to φ even if nobody has the authority to hold them accountable for failing to φ, the OAN is false.Footnote ⁴⁰

It is crucial to note that this argument does not provide a case against the juridical view and its version of the Palsgraf Principle. It only shows that for a juridical realm to be legitimate, contingent social practices must be in place. In this view, the OAN is false because it overstates the connection between obligations and accountability as a necessary conceptual truth. But the truth in the vicinity is that there is a limited domain in which it is good for accountability and obligations to correspond: There are ways of holding one another accountable that it is valuable to be able to perform, but that threaten one another’s freedom if done in the absence of legitimating conditions. We therefore have reason to instantiate those legitimating conditions for at least some part of our interpersonal lives. And there is good reason for those practices to conform to the Palsgraf Principle.

3.A. Wrongs in Deliberation

Cornell’s rejection of the OAN follows most directly from:

Or as he puts the point more simply: “potential wrongings do not provide any reasons” (69). This is a counterintuitive thesis for reasons Cornell himself articulates. Cornell stresses two features of moral life that point toward the common-sense view that wrongs do provide reasons for action. The first is his position that “[o]ne of the remarkable things that we can do as conscious, self-governing, temporally extended agents is think about one normative perspective while we inhabit another” (180). At the moment of action, he tells us, we occupy “a kind of limbo”, suspended between our options and the consequences they might have (87). The second is Cornell’s insistence that “focusing on the wrong that one might commit is a way of drawing our attention to the other person, something all too often missing” (92). It is a way of drawing “attention to the fact that moral obligations are often not just free-floating, but rather they are owed to other persons” (92). Surely, one might think, the benefit of this remarkable intertemporality is that we can try to peek into the future and anticipate the consequences of our actions, including the harms and wrongs they might impose on others. Doing so might seem a reasonable way to take others into account in the way morality requires.

But Cornell denies that wrongs give us any ex ante reasons for action. There is some truth, Cornell concedes, to the idea that thinking about the wrongs we do to others brings them into our deliberation in a meaningful way by giving us reasons to avoid acting in a way that wrongs them. This, he says, is a “poetic truth” that gives us motivating reasons (93). But it does not give us justifying reasons (93). The basis for this is not obvious. After all, in Cornell’s view, other people are the sources of our moral obligations. And if a reason draws our attention to the source of our moral obligations, then it seems like a justifying reason. But Cornell’s view is that other people are the sources of our obligations insofar as they are rightsholders, but not insofar as they stand to be wronged. Focusing on the wrong we might commit has motivational benefits insofar as it draws our attention to other people; not because it does so in a way that puts us directly in touch with our reasons for action, but because it reminds us of other people, who happen also to be sources of our real reasons for action.

Cornell prefers this indirect view of the relationship between wrongs and justifying reasons because he thinks it would be “odd” to deliberate with a view toward what one might later come to regret, including the wrongs one happens to do to others (90). “To think about what wrongs one might commit would be to attend to something epiphenomenal” (87). It would be like going “through life trying to avoid having regrets” (87-88), and “there seems to be something amiss about making a choice merely to avoid regret” (90 [emphasis in original, footnote omitted]). Focusing too much on the wrongs we might do to others would be to act too much like Holmes’s infamous Bad Man, concerned only with liability.Footnote ⁴² The jurors who punish corporate defendants who figure liability as a cost of doing business rightly identify something odious in such behavior.

This argument is not convincing. For one, it is only an argument against treating wrongs as motivating reasons: Cornell’s complaint is that the person who is motivated by avoiding liability has deliberated badly, analogous to the way someone who is motivated by avoiding regret has deliberated badly. But we are looking for an argument against treating wrongs as justifying reasons.Footnote ⁴³ Second, there is some tension: Cornell also insists that it is at least sometimes good to treat wrongs as motivating reasons because it reminds us that morality is about other people. Perhaps his view is that we should sometimes focus on wrongs, but not too often; we need to find the right balance between (a) respecting rights, (b) avoiding wrongs because we are avoiding liability or regret, and (c) avoiding wrongs because they remind us of the moral value of other people. Maybe that is correct. But nothing about the argument explains why wrongs do not provide justifying reasons.

The main argument Cornell provides for the view that wrongs do not provide justifying reasons is his discussion of the cases in Part 1.B above. Those cases are meant to show that we can wrong someone without acting counter to any reasons either because we fail to fulfill a promise that created no reasons for action or because our reasons are perfectly balanced in favor of competing options and the presence of a wrong provides no tie breaker. I will make two observations about that argument. First, I do not share many of Cornell’s intuitions about these cases. For example, in Sexual Promise, it seems to me that A can acquire a duty to have sex with C through her promise. Suppose in the context of a committed relationship between two adults, one says to the other: “I promise to perform sex act x, which I know you like, if you promise to perform sex act y, which I prefer.” These promises create rights and obligations. Of course, the duty to perform can be overridden. In Sexual Promise as described, A’s duty can probably be overridden easily. But that does not mean it is not an obligation or that C lacks a right.

Second, these cases seem to introduce a tension in Cornell’s conception of wrongs.

Cornell suggests that for A to wrong C, it is necessary that A’s action is somehow subpar: They act contrary to the weight of their reasons, even if they do not violate an obligation (210-14).Footnote ⁴⁴ But in the cases Cornell adduces for the argument that wrongs do not provide reasons, he claims to identify wrongs that occur when A acts consistently with the weight of their reasons. This seems true of his account of wicked promises and sexual promises (77-81). If A’s only reason to have sex is that she wants to have sex (79), then if she breaks a sexual promise to C because she does not want to have sex, she does not act contrary to what she has reason to do, because she has no reason to have sex with C. But Cornell maintains that A might still wrong C in that case.

I have suggested that Wrongs Give No Reasons is a very strong thesis and that Cornell’s appeals to cases and to the oddity of deliberating to avoid wrongdoing do not provide an adequate defense.Footnote ⁴⁵ But maybe we can argue that wrongs do not provide justifying reasons for action if we can paint a picture of moral deliberation and moral community that does without wrongs as justifying reasons. I turn to that now.

Cornell’s picture seems to be this: We have the morally interesting capacity to think ahead and consider the possible consequences of our actions for other people. We can predict the possible wrongs we might produce. The revulsion or anticipatory regret it would be appropriate to feel in response to a predicted wrong has psychological force. It reminds us that other people have rights and that those rights give us powerful reasons for action. At the same time, it would be “odd” or “amiss” to focus too much on possible wrongs. We really should just focus on our reasons, which often means focusing on others’ rights and the obligations they produce. After we act—and escape limbo—we will see how the chips fall when we reflect on “material facts about what has happened between us” (200). But we will then be in a wholly new scenario. And whether we have to make amends, ex post, is not determined by whether we acted rightly, ex ante.

It might help to make this more concrete. Return to Good Will Hunting. Hunting is considering staying in his hometown rather than chasing his ambition and fully exploiting his talents. On no plausible moral account would Hunting violate a duty to others by staying home. But Cornell believes Hunting at least plausibly wrongs his friends by doing so. Suppose Hunting agrees with Cornell. He might reflect on the wrongs he predicts he might do, and this might be a useful way of reminding him that his friends are the source of important moral reasons. But if he keeps reflecting, he will realize that in the case at hand those friends do not give him any reasons whatsoever not to waste his talents. He can predict that they will resent him (or, as I think is more plausible, that they will be disappointed, perplexed, envious, perhaps piteous). But it would be a mistake to take this prediction as putting him in touch with justifying reasons. Suppose, then, that Hunting stays home and it turns out that his friends do resent him and press valid complaints against him for having wronged them. Now Hunting is on the hook. Now he is answerable to his friends. They can demand apologies or maybe even ice him out for a while.

And now—despite having possessed the remarkable moral capacity to reflect on future wrongs to his friends and to predict this all would happen—Will Hunting is simply in a new situation and must account for wrongs which he had no direct reason to prevent.

I do not find this to be a plausible picture.Footnote ⁴⁶ Often, in the face of some unexpected harm to others, the virtuous thing to do is to say, “It does not matter how we got here. We need to figure out how we are going to carry on.” This may involve taking accountability for things that are not your fault.Footnote ⁴⁷ It may involve not insisting that you did nothing wrong. It may involve a single-minded focus on repair. In my view, this is the attractive truth that Cornell relies on. But emphasizing it too much yields an inaccurate picture of moral deliberation.

3.B. Moral Community

Perhaps more importantly, Cornell’s picture produces a distorted account of moral community. Part of Cornell’s complaint against the OAN and some rights-based views is that they conflate moral community with a system of rights. Because wrongs are an important part of the way we relate to others, and because wrongs and rights come apart, a system of rights only reflects a partial view of our moral interconnectedness.

Cornell’s alternative picture of moral deliberation anchors his alternative picture of moral community. In the ex ante deliberative perspective, we focus on the rights of others and the obligations those rights ground. But this, in Cornell’s view, is an incomplete, heart-pinching way to see moral life. How do we get the rest of it into view? Not by focusing on the wrongs we might to do others while deliberating ex ante. We act and then we wait and observe. What turns out to sever the connection between rights and obligations, on one side, and wrongs and accountability relations, on the other, is time and the contingency it reveals. It is then time that brings us back together again by revealing the wrongs that our actions turned out to cause. The full picture of our interconnectedness comes into view when we, ex post, survey the terrain and repair the wrongs we turn out to have done.

I doubt that we fully realize the extent of our interconnectedness only when we wrong one another. Here is another way to think about the link between moral community and accountability. Above I suggested that A can have a moral obligation to φ even if nobody has the authority to hold them accountable for failing to φ. This will be the case when the legitimating conditions for accountability are absent because, for instance, we have not developed the appropriate social practices. Although this is a familiar enough phenomenon, it signals an unstable moral situation. It is good when victims have the authority to hold wrongdoers accountable. Doing so is an important way of asserting one’s dignity and is an important component of self-respect.Footnote ⁴⁸ This implies that when wrongs and rights come apart due to a deficit in authority, we fail to live in a fully-developed moral community. That in turn suggests a practical response: We must figure out what it takes to constitute the authority needed to hold one another accountable. This could involve either reducing reasonable disagreement or developing institutions or social practices to work through reasonable disagreement. On this picture, the fullness of our interpersonal life is not something revealed to us through the wrongs that unfold in time. It is something we reveal and build through our active cooperation to develop and sustain social practices. I submit that this is a more attractive picture and that it falls out of the kind of juridical approach I have suggested: one that embraces the Palsgraf Principle but denies the OAN.