RESEARCH · DEVELOPMENTAL READING EDITION
Chapter IV — The Normative Life of an Ending
Why legitimate endings can leave responsibilities, and why regret is not always evidence of wrongdoing.
← Complete contents · Chapter 4 of 8 · Updated November 2026
In this chapter
- 4.1. A permissible decision and an unfinished world
- 4.2. The permission to exit is not a universal power of release
- 4.3. Betzler's ethics of divorce: the strongest direct predecessor
- 4.4. Lopez-Cantero and the limits of demanded deliberation
- 4.5. The non-substitutability of intimacy and the substitutability of some obligations
- 4.6. Regret without wrongdoing: the moral remainder after justified action
- 4.7. A taxonomy of what may remain
- 4.8. Character change and the survival of historical responsibility
- 4.9. The caregiving handoff and the difference between help and obligation
- 4.10. Safety, coercion, and the duty that must not be imposed
- 4.11. Multiple claimants and incompatible but legitimate interests
- 4.12. A transition without wrongdoing: the hardest test
- 4.13. Five statuses of an obligation at the moment of change
- 4.14. A two-sided constraint on the terms of departure
- 4.15. Rival theories and the originality burden
- 4.16. The place of this chapter in the dissertation
- 4.17. Conclusion: ending is an act, not an erasure
4.1. A permissible decision and an unfinished world
The relationship ends before the shared life has finished depending on it. When Adrian tells Morgan that he wishes to separate, he does not confess an affair, announce an intention to punish, or claim that Morgan has done anything wrong. The love that once gave their partnership its distinctive form has changed. For several years the couple shared a home and maintained an arrangement under which Morgan worked fewer paid hours while Adrian pursued a demanding career. Morgan's career opportunities declined, but household income increased, and the couple jointly treated the sacrifice as worthwhile. Adrian now wishes to live alone. He is willing to discuss property, financial arrangements, and the logistics of the separation, but he cannot promise to continue the emotional or romantic involvement that had previously made the partnership intelligible to both of them.
Morgan objects that Adrian's freedom to leave cannot erase the life they constructed. The objection is serious. A person who shares a life with another can make assurances, invite sacrifice, and influence what opportunities remain credible or practical. When the relationship ends, those effects do not disappear because one person's sentiments change. Yet the conclusion Morgan might most desperately want—that Adrian must continue the relationship—does not follow from the legitimacy of the grievance. Something may be owed even where its original form cannot be demanded; conversely, the fact that the relationship was once valuable does not imply that any specific remedy is required. The proper moral question begins where the binary alternatives of abandonment and captivity fail.
Suppose Adrian acted neither dishonestly nor callously in reaching his decision. Assume he gives adequate notice, communicates the fact of his changed commitment without manufacturing allegations, and accepts that previously undertaken financial duties may require continued performance. Assume also that Morgan's losses are genuine: reduced earning capacity, disrupted plans, and the collapse of a future they had jointly imagined. The situation may contain a permissible departure, legitimate hardship, and continuing obligations without requiring us to describe the departure itself as wrongful. We must ask, instead, which reasons survive the relationship's transformation and what they authorize each party to demand.
The case is not designed to idealize Adrian's decision or to minimize Morgan's predicament. It is designed to make their claims analytically separable. A right to end romantic participation would have little moral meaning if another person's reasonable expectation of continued love automatically defeated it. But that right would be ethically inadequate if invoked to negate every responsibility arising from a history of commitments and shared decisions. The difficulty is to determine the extent and objects of continuing responsibility without silently converting responsibility for consequences into an entitlement to control the former partner's future.
The earlier chapters established two distinctions: a need is not the same as a duty owed by a particular person, and performing a task is not the same as having authority to undertake or discharge that duty. We then examined the promises, reliance, shared commitments, vulnerability, and care that may make obligations binding. Now those grounds face change. What remains when a relationship ends? What can properly expire? What requires repair? And when does a demand for continuing responsibility become an attempt to prevent a legitimate departure? An ending can raise all these questions even when leaving is not itself wrong.
4.2. The permission to exit is not a universal power of release
A permission to cease an act and the power to extinguish an obligation are different things. A person may be permitted to resign from a job without acquiring the power to erase debts incurred during employment. A clinician may be entitled to retire without being authorized to disclose patient information indiscriminately or disregard particular duties concerning records. A person may end an intimate partnership while still owing truthful communication, agreed financial support, or the return of property. These are not paradoxes. They concern duties with different objects and sources, and the permission applicable to one object does not determine their fate.
The distinction can be made through the Hohfeldian categories introduced in Chapter II. A liberty to withdraw from one form of participation is not automatically a power to cancel all other directed duties. Nor does another person's possession of a claim to some performance imply a power to prevent every action that would affect that claim. We need to identify which normative incidents are present, which are protected by independent duties, and which can legitimately be altered through waiver or agreement (Wenar, “Rights”). The conceptual taxonomy does not settle the substantive moral issue, but it prevents the purported right to depart from being used as a rhetorical solvent.
Consider a professional who has promised to complete a specified project, but later becomes medically unable to perform the promised work. Permission to cease the unsafe activity may be well founded. Whether any duties remain concerning notice, records, reimbursement, cooperation, or previous wrongdoing requires further analysis. Some may have been explicitly undertaken; others might follow from applicable professional roles or standards of due care. Yet the very same facts may support a judgment that direct continuation is not owed. To say “the obligation persists” without specifying which obligation is therefore misleading. The past can retain a claim upon some acts while losing its claim upon others.
The inverse mistake belongs to the beneficiary. Suppose Morgan has a valid financial claim arising from the couple's joint arrangements. The claim does not by itself provide authority to determine whom Adrian may love, whether he may live alone, or whether he must remain available for conversations until Morgan judges the decision sufficiently explained. There may be separate duties of communication and transition; those duties require their own grounds and limits. The right to demand one form of performance cannot be presumed to encompass all conduct that would ease the recipient's loss.
This analysis is already extensively anticipated by philosophical accounts of normative powers, promising, and relational freedom. The significance of the chapter is not that it discovers the difference between permission and release. It is that it forces the difference to be made explicit when historical intimacy gives moral claims an emotional intensity that can conceal their objects and limits. A theory of relational transition must be able to recognize that departure is permissible without pretending it is inconsequential, and to affirm a real obligation without pretending it entitles the claimant to comprehensive jurisdiction over another person's future.
4.3. Betzler's ethics of divorce: the strongest direct predecessor
Monika Betzler's “Ending a Special Relationship: Toward an Ethics of Divorce,” published in 2024, makes it untenable to present the general problem of permissible separation and surviving duties as unexplored. Her central proposal gives relational freedom an enabling role: the possibility of divorce is not simply a regrettable concession to failed relationships but a condition of the distinctive value of romantic partnership. A relationship whose continuation must be compelled would not embody the same kind of relational good. Yet those ending a partnership can wrong one another in distinctive ways. Betzler analyzes claims to explanation, participation, and transitional goods alongside the initiating partner's right to relational freedom, describing the corresponding duties as residual duties grounded in the prior special relationship (Betzler 2024).
The philosophical importance of the account lies in its rejection of both extremes that our opening case might tempt. The partner who leaves is not automatically a wrongdoer; the partner left behind is not automatically a person with no remaining claims. A special relationship can generate duties whose force outlives the relationship's ordinary form. The vocabulary of residual duties is therefore not a discovery to which this dissertation is entitled. It has a contemporary, detailed, and directly relevant predecessor.
Indeed, Betzler's subsequent “Post-Divorce Duties: What We Owe Ourselves and Our Exes as Former Partners,” published in 2025, extends the inquiry beyond the moment of separation. The chapter identifies duties concerning the confidentiality of entrusted information, letting go, grieving, and decency as important ethical features of former-partner relations. These duties need not be directed solely toward restoring benefits once enjoyed; they can govern how former partners treat one another after the relationship has ended and how they relate to their own continuing lives (Betzler 2025). This is a more far-reaching account than a minimalist obligation to settle financial affairs. It considers the relation that remains after the romantic relation ceases.
Some proposed duties after a breakup require especially careful interpretation. A duty to grieve, if it exists, cannot simply mean that a former partner may command someone to feel or display sorrow. A duty to let go may protect the other's freedom to build an independent life without giving either party authority to dictate the other's pace of recovery. The available abstract names these duties but does not establish their complete scope or conditions. That limit matters. The relevant scholarship already offers a demanding ethics of relationship endings; before arguing for something new, we must examine its full defense.
Betzler's account offers a powerful explanation of Morgan's position. Morgan may have a legitimate claim to a fair settlement and to some forms of respectful participation in practical decisions affecting their shared history. Adrian's permission to withdraw from romance does not settle those claims. At the same time, Betzler's invocation of relational freedom offers reasons to reject an unlimited demand for personal continuation. The outstanding question is not whether obligations can survive separation, but what justifies particular obligations when they conflict with the conditions that make a genuinely voluntary relationship valuable.
A charitable reconstruction must not treat Betzler as proposing that an unwilling person can be required to remain romantically committed until a partner agrees. Her argument is expressly structured around both kinds of rights. The pressure arises when the content of explanation or participation duties is unspecified, because a duty that appears modest in abstract formulation may become burdensome, indeterminate, or controlling in circumstances of profound disagreement. That problem has already been identified in the literature; the next section examines its most direct critique.
4.4. Lopez-Cantero and the limits of demanded deliberation
Pilar Lopez-Cantero's 2026 response to Betzler challenges what she calls deliberative break-up duties. She argues that demands to explain and facilitate participation can be insufficiently action-guiding where the departing partner's explanation is simply that romantic love has ended. Even a weaker duty to make an effort to deliberate, she contends, carries moral risks serious enough to warrant rejection. Her criticism is not an endorsement of callousness; it questions whether imposing particular forms of rational exchange captures the relevant moral meaning of relationship dissolution (Lopez-Cantero 2026). The disagreement is substantive and current, and our dissertation must work within it rather than claim to have discovered the concern independently.
Imagine Adrian tells Morgan, accurately, that he no longer wishes to remain in the relationship and cannot offer an explanatory narrative that Morgan will find satisfying. Morgan replies that the relationship cannot legitimately end until Adrian provides reasons that can be evaluated and jointly discussed. Such a demand might appear respectful because it invites deliberation rather than physical coercion. Yet it can locate effective authority over departure in the person who may always reject the explanation as insufficient. What begins as a claim to be heard becomes, without clear limits, a power to delay or condition the exercise of relational freedom.
The objection does not mean that concealment, deliberate misrepresentation, or sudden abandonment is always permissible. A person who has lied about finances or arrangements may owe truthful disclosure. Someone who has induced serious reliance may owe notice. A former partner may have standing to participate in decisions about shared property, dependents, or contractual obligations. But such duties concern specified matters. A duty to give information necessary for practical coordination is not automatically a duty to expose one's emotional life to adversarial examination. The objects differ, and the claimant's authority must be justified separately.
Nor should communication be reduced to the exchange of propositions. A person may convey care through acknowledgment, gestures, logistics, or a willingness to hear the consequences of departure without conceding that every feeling requires a rational defense. Lopez-Cantero's analysis presses the possibility that a deliberative model can privilege a narrow conception of reasons where the end of romantic attachment is not fully susceptible to argumentative reconstruction. The character of the underlying relationship may make a demand for exhaustive explanation inappropriate even when ordinary truthfulness remains owed.
Neither an unconditional demand for explanation nor an unconditional right to silence is adequate. Some responsibilities to communicate are consistent with leaving; others effectively make departure depend on the former partner's permission. The difference may turn on the risks of contact, the costs of silence, earlier promises, practical entanglement, and the claim being made. Betzler and Lopez-Cantero have already examined this dispute in substantial detail. Any further proposal must offer a better-grounded limit or a genuinely different explanation, not simply restate the opposition between accountability and freedom.
4.5. The non-substitutability of intimacy and the substitutability of some obligations
One might attempt to resolve the dispute by dividing all relational obligations into two groups: those that may be satisfied through substitutes and those that must be performed personally. Romantic participation seems an obvious candidate for the second group, financial support for the first. Yet the distinction does not yield a complete theory. A person may not be morally entitled to compel anyone's romantic love even when another person cannot supply an adequate substitute. Conversely, a promise of personal presence can sometimes be transformed by agreement, while a financial debt may remain directed to a particular creditor despite the availability of other recipients who would value the payment more.
The difficulty is not only one of fungibility. It concerns whether the original object of a duty can legitimately be demanded. If Adrian no longer wishes to continue a romantic partnership, Morgan's suffering cannot by itself make compelled romantic presence a permissible solution. An attempt to reproduce the relationship through duty alone risks hollowing out precisely the voluntary attachment that gave it value. But if the couple agreed to a property distribution or other financial protection, the failure of the relationship does not make those commitments disappear. The persistence of the financial claim is not equivalent to an entitlement to restore the intimacy that once accompanied it.
Nor can Adrian automatically choose a convenient substitute for every unfinished responsibility. Suppose Adrian humiliates Morgan publicly during the separation and later sends a substantial payment in lieu of acknowledgment. Money may be valuable and might discharge an independent financial duty, but it need not constitute an adequate response to an injury to dignity or reputation. The same is true of a gift offered instead of returning confidential information or correcting a serious factual misrepresentation. An account of moral repair must attend to what kind of wrong occurred and the particular standing of the person wronged; the successful provision of a benefit does not, by itself, establish discharge.
The question therefore contains two separate normative limits. The person owed a duty does not automatically control every acceptable substitute, and the person owing the duty does not automatically get to select whichever substitute is easiest. These limits are already anticipated in the literature on normative powers, directed duties, and moral repair, including Shiffrin's analysis of hidden delegations and Räty's discussion of the victim's qualified prerogative over amends (Shiffrin 2023; Räty 2026). Their relevance to endings is clear, but their theoretical priority must be acknowledged.
The constructive challenge is to identify when an obligation's original performance has become impermissible, when the obligation can be satisfied through an acceptable alternative, and when a different duty survives because the earlier action generated a distinct claim. These events may coincide in practical experience, yet they are not the same normative transformation. Chapter V will focus on the rules governing their attribution and discharge. For now, the important point is that an ending can bring genuine discontinuity of intimate participation alongside continuity of other, differently grounded responsibilities.
4.6. Regret without wrongdoing: the moral remainder after justified action
Not every morally significant remainder is a surviving obligation, and not every justified action leaves the world morally untroubled. A battlefield medic may cause grave pain in the only available attempt to save a patient's life; a person may prioritize one urgent duty while another person's interests suffer; a caregiver may stop performing unsafe work knowing that the beneficiary's immediate prospects will deteriorate. The person acting may appropriately experience regret without being properly understood as having committed a wrong. The distinction between regret and guilt is central to disputes concerning moral dilemmas and moral residue. The Stanford Encyclopedia's account of moral dilemmas emphasizes that regret may be fitting even where the agent has not done anything all-things-considered wrong. An appeal to negative emotion alone therefore cannot establish that a genuine moral requirement was violated (“Moral Dilemmas,” §6).
This matters because ordinary judgments about endings often treat distress as though it must originate in another's culpability. Adrian feels sorrow at Morgan's loss; Morgan experiences the sorrow as confirmation that Adrian recognizes a wrongdoing. Yet Adrian's regret may concern the fact that a permissible decision has painful consequences. Likewise, a caregiver's sadness at a dependent person's difficulty does not establish that the caregiver wrongfully ceased. The emotion can reflect an adequate responsiveness to others' lives without acknowledging that any specific directed duty was breached.
The conclusion does not render the feeling ethically empty. Appropriate regret may motivate voluntary help, better communication, efforts to ease a transition, or attention to public injustices that made the loss severe. Those responses can be admirable, and sometimes independent duties of beneficence or care make a particular response obligatory. But such further duties require grounds. The reason it is appropriate to regret an outcome may not be the reason one is required to compensate for it. A theory that treats regret as a proxy for liability risks transforming moral sensitivity into evidence against the person who displays it.
The distinction becomes especially difficult when the agent knowingly causes a loss. Adrian's departure may predictably worsen Morgan's economic position. Mara's justified withdrawal may predictably unsettle a dependent person. Causation is morally relevant, particularly where an agent induced reliance or voluntarily undertook a protective role. But causation alone does not establish wrongful causation, nor does it always specify a remedial duty. One can contribute to an adverse outcome through a permissible act, and the fact of contribution must be interpreted against the background of commitments, fair expectations, alternatives, and competing responsibilities.
Philosophers disagree about what remains when legitimate choices sacrifice something morally important. Some hold that conflicting duties can leave moral failure whichever path is chosen. Others distinguish defeated preliminary reasons, duties that still bind, and regret appropriate to goods that could not be preserved. This chapter need not resolve the entire dispute. It must, however, resist two inferences: regret alone does not prove that someone was wronged, and a justified decision does not prove that no claim remains. What follows an ending might be a duty, a reason to acknowledge loss, an appropriate emotion, or a claim requiring separate justification.
Consider two variants of the opening case. In the first, Adrian and Morgan had made no relevant financial agreement, both maintained equivalent earning capacity, and the separation entails profound emotional loss but little induced material reliance. In the second, Adrian explicitly invited Morgan to abandon a professional opportunity for a jointly planned future while assuring Morgan of long-term financial protection. The same loss of romantic attachment might justify separation in both cases. Yet the second history contains a potential ground for continuing financial obligations absent from the first. The difference is not how regretful Adrian feels. It is what the parties legitimately undertook and how that undertaking bears upon the consequences of departure.
4.7. A taxonomy of what may remain
An ending may leave at least five different kinds of moral remainder. An earlier duty may still bind, as when someone must return property or make a promised payment. The ending may create or modify a duty: reasonable notice, a proportionate handoff, or correction of misinformation. A separate duty to repair wrongdoing may survive, such as compensation for a broken promise. In other cases, what remains is a reason for regret, sympathy, or voluntary support rather than something the other person can demand. Finally, a need may call for public or collective provision without giving the former provider an indefinite personal obligation. These possibilities can overlap, and the list need not be exhaustive; the important point is that loss, wrongdoing, duty, and need are not synonyms.
The distinctions matter because each kind of remainder has different conditions of satisfaction. Payment can discharge a valid monetary obligation, but may not discharge a duty of confidentiality. An apology may address a wrong without restoring a lost service. Regret may be appropriate without entitling another to a settlement. A public agency may owe replacement care even when the departing caregiver has no continuing duty to arrange it. Calling all of these “remaining obligations” hides the grounds, bearers, and objects that give the category moral significance.
The third category is particularly liable to distort the analysis of morally permissible departures. If no wrong has been committed, it would be mistaken to describe all transition assistance as restitution. Yet the absence of a wrong does not prove that duties of notice or cooperation are absent, since those duties may have been independently undertaken or may arise from legitimate roles. A clinician's responsibility to provide essential information for continuity of care need not presuppose that retirement is wrongful. A parent who arranges competent replacement care need not be compensating a child for the wrong of changing living arrangements. The language of repair, while powerful, should not colonize every ethical transition.
The fourth category is often neglected because it does not yield an immediately enforceable conclusion. A former friend may have reasons to hope for the other person's well-being without owing renewed friendship. Someone ending a volunteer role may appropriately recognize the effects of departure without being bound to preserve an organization's previous level of service. Such reasons can be morally important even when no recipient can insist upon their particular satisfaction. To deny them because they cannot be translated into directed claims would impoverish the ethics of relationships. To translate them automatically into claims would turn their moral significance into coercive leverage.
The fifth category requires special attention to background injustice. When a private arrangement has long compensated for inadequate institutional care, the withdrawal of one individual may expose a need that the institution already ought to have addressed. Ong and Savulescu's account of relational responsibility explicitly broadens the moral field beyond individual dyads, recognizing collective responsibilities to mitigate excessive caregiving burdens (Ong and Savulescu 2026). A transition may accordingly reveal a moral problem that predates the departing agent's decision. In such cases, the apparent demand for personal continuity can function as a substitute for confronting the real distribution of responsibility.
The taxonomy is a working analytical instrument. Its elements have substantial predecessors in the literature on promises, moral repair, special duties, beneficence, and institutional justice. Its immediate purpose is to prevent one form of remainder from being mistaken for another. Whether the dissertation can show that an additional category or principle is necessary remains an open research question.
4.8. Character change and the survival of historical responsibility
Adrian's loss of love does not ordinarily raise a mystery about whether he is the same person. He made the earlier commitments, and changed feelings alone do not show that any surviving duty has expired. More radical changes in personal identity raise a harder question: can obligations endure when psychological continuity is seriously disrupted? Søren Holm's “Moral Obligation and Changes in Personal Identity” (2011) defends the survival of many relational obligations across such changes and considers parallels with legal personality and state succession (Holm 2011). The idea that past commitments can bind a profoundly changed person is therefore not untouched philosophical territory.
This precedent prevents two facile arguments. One cannot claim that because a person has become different, earlier obligations naturally evaporate. Nor can one claim that the persistence of every earlier duty follows simply from numerical identity. The specific relation, content, and legitimate conditions of the obligation remain important. A promise to repay money may survive substantial changes in religious belief or professional calling. A promise to sustain voluntary romantic attachment is not straightforwardly a guarantee that one's future desires can be controlled. What survives must be determined according to the grounds of the undertaking, its limits, and the relevant rights of all parties.
It is also possible for an agent's changed capacities to affect what can be required without changing the moral value of what was originally promised. A caregiver who develops a disabling illness has not necessarily repudiated the beneficiary's importance. The ethical question concerns the object of the duty and the feasibility or permissibility of its performance under altered circumstances. A theory that equates withdrawal with diminished love or bad character may obscure the material conditions of responsible action. A theory that treats every hardship as extinguishing prior undertakings may fail to protect legitimate reliance.
This distinction touches the author's existing work in Do Not Flatten Me, which argues against granting past descriptions unlimited jurisdiction over a developing person, and After the Gift, which rejects possession of another person's future through generosity. Those works supply conceptual ancestry, but the present inquiry concerns a more exact normative issue: which historically generated claims remain valid, which acts of accountability are owed, and who can change or discharge those claims. Without this precision, the dissertation would repeat its author's earlier defense of becoming in the language of obligation.
Holm's argument also warns against exaggerating the philosophical consequences of ordinary personality change. Not every shift in ambition or attachment constitutes the sort of personal-identity change that drives metaphysical debates. Our principal cases do not require settling that debate. They require explaining why legitimate claims may persist or change even when the responsible human being is plainly the same person and why changed circumstances can affect duties without erasing historical answerability.
4.9. The caregiving handoff and the difference between help and obligation
Return to Mara, Leila, and Clara from Chapter I. Mara can no longer safely provide direct care. Clara needs assistance, Leila depends on an arrangement that has stabilized the household, and a professional caregiver could assume some of the tasks. Does Mara owe a handoff? The temptation to answer yes is understandable: care transitions can be dangerous when information about routines, medications, and risks is lost. But a moral justification must do more than observe that a handoff would help. It must show that Mara has a particular obligation to facilitate it, and establish what action is proportionate to her circumstances.
A promise may provide one ground. If Mara expressly undertook to provide a specified period of notice and to document essential routines, that commitment has force unless defeated by sufficiently weighty circumstances. Induced reliance may provide another. If Mara knowingly discouraged alternatives and gave Leila reason to expect continuity, a duty to correct the expectation may arise even absent a formal promise. An accepted formal role may provide a third ground, especially where public standards establish duties of safe transition. General beneficence may also require limited emergency action where the cost is small and the threatened harm severe. These are distinct routes to duty, and their existence should not be inferred from the fact of Clara's continued need alone.
Now suppose Mara offered help freely, repeatedly refused a permanent commitment, and warned Leila months earlier that she might have to stop. Leila nevertheless postponed arranging services because Mara was exceptionally competent and her assistance was free. If the resulting need for a replacement creates an unlimited duty for Mara to train one, Leila's unilateral planning decision becomes a power to extend Mara's obligations. That outcome would threaten the distinction between justifiable reliance and institutionally convenient dependence. It would also risk converting the fact of Mara's generosity into evidence that she has assumed responsibility for others' failure to prepare.
The opposite extreme is equally indefensible. Suppose Mara alone knows where Clara's life-saving medication is stored and could communicate the information safely in a minute. A refusal to give the information might be wrongful under an appropriate principle of rescue or due care, even if Mara is entitled to leave the broader caregiving role. But this duty is not an indefinite duty of care. Its object is a feasible act addressing a concrete risk. The case demonstrates why an obligation can remain when the original relationship legitimately ends without becoming a demand to maintain the original arrangement.
We must also consider Clara as an independent moral participant, not a passive object of negotiations between Mara and Leila. Clara may have preferences concerning caregivers, confidentiality, living arrangements, and the acceptable burdens of transition. Depending on capacity and the institutional setting, Clara's wishes may be relevant to authorization, refusal, and the legitimacy of proposed substitutes. No theory should presume that Leila automatically possesses unrestricted authority to decide for Clara or that Mara's departure licenses a transfer of confidential information without appropriate safeguards. A genuine relational account must include the standing of the person receiving care.
The handoff case is therefore not a shortcut to a universal transition duty. It is a laboratory for distinguishing independent grounds of limited assistance from claims of indefinite personal liability. Kittay and Tronto's care ethics, Chadha-Sridhar's grounding theory, and Ong and Savulescu's network account already supply many of the necessary resources. The dissertation's task is to determine whether they leave an important normative event unspecified when the role ends and another person's need remains.
4.10. Safety, coercion, and the duty that must not be imposed
The ethical language of responsible exit becomes dangerous when it is used to require sustained engagement with someone who exercises coercive control. Imagine a person leaving a relationship characterized by threats and surveillance. The controlling partner claims that departure without a face-to-face discussion is unfair, that the relationship's history creates a duty of explanation, and that a transition cannot be legitimate until both parties have agreed upon its terms. The demand has the language of procedural fairness while functioning as a mechanism for continued access. A theory that universally requires dialogue, notice, or negotiated release would be morally deficient in such a case.
The objection is not resolved by declaring that the departing person owes nothing under any circumstances. There may be children, shared property, contractual obligations, or other protected interests requiring safe forms of coordination. But the means of meeting those responsibilities must not be identified with interpersonal availability to the abusive party. Legal representatives, mediated exchanges, emergency services, or protected communication channels may be appropriate in particular circumstances. The philosophical point is that a duty's existence does not determine that its fulfillment must take a form exposing the duty-bearer to unacceptable danger.
A second coercion problem can arise without physical danger. A person might be compelled to explain an intimate decision repeatedly under the threat of reputational ruin or economic retaliation. If the burden of showing that one's departure is sufficiently justified never ends, the purported freedom to leave becomes dependent upon another person's discretion. Even a formally modest duty of deliberation can be converted into a power to delay closure. Lopez-Cantero's challenge to Betzler is particularly important here because it asks whether some proposed duties are action-guiding and what moral risks their weaker versions carry (Lopez-Cantero 2026).
A third problem concerns institutional roles. A worker may be asked to stay until a replacement is trained, yet the employer alone determines whether a replacement is hired. The employer can thus extend the purported transition indefinitely by refusing to fulfill its own planning responsibilities. A moral principle that automatically assigns the costs of continuity to the departing worker would reproduce a structural form of coercive dependence. The fact that the resulting service loss is foreseeable does not establish that the worker must supply the remedy.
These cases demand an anti-coercion constraint: an asserted residual duty should not be specified or enforced in a way that grants an affected party illegitimate control over the person whose original participation has permissibly ended. But this constraint is not enough to determine positive obligations. It must operate alongside independently grounded duties and the moral standing of other affected persons. Its role is to exclude impermissible means of securing legitimate goods, not to treat all demands associated with an ending as suspect.
4.11. Multiple claimants and incompatible but legitimate interests
Even a separation involving two partners rarely affects only those two people. Children, dependents, collaborators, employers, and institutions may have built their plans around the arrangement. A departure may be legitimate while still affecting claims that arise independently of the former partners' relationship. Asking only whether Adrian owes Morgan continued participation could overlook a child's welfare or another person's contractual rights. Likewise, Mara's withdrawal affects Clara's care, Leila's work, and institutions that may have failed to provide alternatives. Each affected person may have a different claim grounded in a different source.
Suppose Adrian and Morgan have a child. Adrian remains entitled to end the romantic relationship; the child's claim to appropriate support is not therefore diminished by the couple's loss of affection. Indeed, the obligations to the child may be differently structured from obligations between former partners. The child did not undertake the romantic commitment and should not be made responsible for its failure. An account that treats all post-separation duties as moral residue of the adult partnership would misdescribe the independent grounds of parental responsibility.
Now add Morgan's elderly parent, to whom the couple has provided assistance. If the arrangement rested on a genuine promise by Adrian, some duty may survive separation; if it rested on an informal pattern of mutual help, the grounds require separate assessment. The parent may have rights to essential care that do not depend on Adrian's personal commitment, and public institutions may bear duties to make services available. There is no guarantee that the person bearing the greatest practical impact is the person with the strongest normative claim against Adrian. The comparison requires more than ranking the intensity of suffering.
The plurality of claimants also limits the apparent power to release duties. Morgan might waive a personal right to financial assistance; Morgan cannot necessarily waive a duty Adrian owes their child. A caregiver may be authorized to share specified information with a successor but cannot presumptively release a hospital from its responsibilities. The question of who may modify a duty depends on what that duty protects and how its authority is constituted. No single participant's acceptance of a settlement therefore proves that all obligations arising from a shared arrangement have been discharged.
What the cases reveal is an asymmetry between the decisiveness of the exit choice and the complexity of its consequences. A particular person may have legitimate authority to end personal participation even when others retain substantial interests that must be addressed through different mechanisms. This asymmetry is one of the reasons a theory of temporal jurisdiction is attractive. Yet as a general insight it can be reconstructed from established theories of rights, role-based obligations, care, and institutional justice. The dissertation must still show what additional explanation it supplies.
4.12. A transition without wrongdoing: the hardest test
The strongest case for this dissertation is not a person who violates a clear promise and then asks to be forgiven. Ordinary theories of breach and repair already explain much of that situation. Nor is it a person who simply rejects a duty whose scope is settled by legitimate law or institutional role. The difficult case is a morally permissible transition that imposes serious losses without the existence of a single culpable party and without an uncontested method for assigning replacement responsibilities.
Imagine an informal arrangement in which three adults, Nia, Tomas, and Grace, have for years coordinated care for a fourth person, Salma. They did not enter a comprehensive contract. Each undertook tasks compatible with their employment and capacities, and their coordination was largely competent. The arrangement developed through mutual trust and accumulated judgment. A new disability makes it unsafe for Nia to continue the overnight work that she has usually supplied. Tomas can provide some daytime support but cannot reduce his hours without losing essential income. Grace lives farther away and can contribute money, but no longer has the health required for direct care. Salma wishes to remain at home and opposes a particular institutional placement; alternatives exist but require time and resources. No person in this scenario need be dishonest, blameworthy, or indifferent. The care arrangement nevertheless cannot continue as before.
The question “Who ought to take over Nia's nights?” appears urgent and concrete. Yet its grammatical form may presuppose a single successor to a duty that was never held in that form by one person. Perhaps Nia owed reliable performance under a finite arrangement and has now fulfilled it. Perhaps she owes notice and transfer of information while the duties to secure adequate care are independently shared among family and public providers. Perhaps Grace's financial capacity gives her stronger remedial responsibility than Tomas's ability to perform the task, or perhaps Salma's preferences and rights constrain otherwise efficient alternatives. The case does not invite the fiction that all burdens can be made costless. It demands a justification for assigning each burden.
Suppose Nia provides a month of notice and a detailed care plan, while Tomas and Grace argue that no solution should be implemented unless Nia personally supervises the replacement for six additional months. Their demand could be reasonable under a specific promise of supervision; it could also impose an unjust extension of the role she is entitled to end. The fact that Nia has the most experience is a relevant practical consideration, but the moral force of experience depends on its relation to the alleged duty. Expertise can make an act efficient without making every efficient act owed. In a society that treats competence as a perpetual source of liability, the most capable helpers would acquire expanding obligations precisely because they have already contributed successfully.
Now suppose Nia refuses even a brief transfer of knowledge she can provide safely, and the refusal predictably creates acute avoidable danger for Salma. That may activate general duties or narrower obligations arising from the prior arrangement. This judgment need not imply an obligation to continue overnight care or to guarantee that the entire replacement system will succeed. It identifies the act in question and asks for the normative grounds requiring it. The difference is central: duties should be specified in terms of justified acts and protected interests rather than inferred from an aspiration to avoid all losses associated with transition.
What is original here? Not the recognition that multiple persons have distinct claims, that duties can survive withdrawal, or that institutions may bear obligations. The best combined theories of care, promising, rights, and remedial responsibility already offer these possibilities. The case will be philosophically useful only if it exposes a discriminating judgment or structural inconsistency that those theories cannot adequately settle. At present, it provides a demanding test case, not a proven gap in the literature.
4.13. Five statuses of an obligation at the moment of change
For purposes of the argument that follows, an obligation may occupy several distinct normative positions at the time a relationship changes. An original duty may remain in force, even if the form of ordinary interaction surrounding it disappears. A duty may be fulfilled by an authorized act satisfying its content. A duty may be validly altered or waived by someone possessing the relevant normative power. A duty may terminate because the conditions grounding it no longer obtain, without implying that all needs formerly addressed by it have vanished. And a different duty may arise against another party on grounds that are not identical to those of the original duty. These possibilities must not be mistaken for interchangeable descriptions of one event.
We can make a modest formal distinction. Let D(A,B,φ,t) represent a directed obligation owed by A to B concerning performance φ at time t. Let E(A,R,t₁) indicate that A is permitted to end form R of participation at t₁. The following implication is invalid: E(A,R,t₁) ⇒ ¬∃φ D(A,B,φ,t₁). Permission to end the original participation does not establish the absence of all directed duties. But the converse implication is also invalid: if B has an unmet need at t₁, then D(A,B,φ,t₁) necessarily persists for the act A previously performed. A legitimate need cannot, on its own, specify its guarantor.
We should also reject an inference from appropriate negative emotion to breach. Let G(A,x) denote that A has fitting regret about consequence x, and W(A,B,ψ) denote that A wronged B by failing to perform ψ. G(A,x) does not imply W(A,B,ψ). A justified choice can cause serious loss and elicit regret without the agent having violated the directed duty in question. Equally, the absence of regret does not prove that the duty was fulfilled. Normative facts are not mechanically determined by the emotional life of the actor.
A further distinction concerns the power to demand alteration. If B holds a claim D(A,B,φ,t), that does not by itself establish that B has power to substitute ψ for φ, or to prolong the relationship until ψ can be delivered. Nor does A's permission to end R establish that A has the power to discharge D by unilaterally supplying something else. The justification for alteration must come from the duty's original terms, a valid waiver or agreement, a relevant normative principle, or some other defensible ground. That requirement follows naturally from existing accounts of Hohfeldian powers and promissory authority; it is not a new theorem.
The formalism is deliberately minimal because the moral question is substantive. An obligation can be difficult to interpret without being indeterminate in every respect. An agent may face several competing duties that cannot all be performed. A beneficiary can possess urgent needs without holding an exact claim against an identifiable individual. A legal permission to leave may coexist with a moral reason to assist. The notation is useful only insofar as it prevents a conclusion concerning one of these relations from silently replacing a conclusion concerning another.
4.14. A two-sided constraint on the terms of departure
The preceding analysis suggests a constraint that is plausible but not independently original: the person who leaves cannot treat permission to withdraw as a universal release, and the person affected cannot treat historical dependence as a universal power to require continued participation. Both parties must justify claims by reference to their proper objects and grounds. The constraint is two-sided because an ethical theory must guard against opportunistic escape from genuine commitments as well as coercive extension of duties that were never undertaken.
The first danger appears where a powerful actor exits after arranging an unequal distribution of costs. Suppose Adrian encouraged Morgan to relinquish professional opportunities under an explicit long-term financial understanding, then invokes the sincerity of his changed emotions as proof that the past has no claim upon his resources. This would mistake a legitimate liberty concerning intimacy for a power to terminate independently grounded financial duties. The fact that love has ended does not undo the history of assurances. The moral relevance of the earlier undertaking is not a disguised demand for romantic reconnection.
The second danger appears where an affected person uses a legitimate material claim to require something whose moral value depends upon free participation. Suppose Morgan is entitled to financial support but insists that Adrian must also continue attending intimate family gatherings as a romantic partner until the financial situation is entirely restored. The financial obligation cannot automatically justify that expansion. The legitimacy of support is not a title to govern another's emotional self-presentation. A separate practice of coordination might justify particular conduct in specific circumstances, but such a claim must stand or fall on its own grounds.
The third danger involves third parties. A couple may agree on the disposition of shared resources yet purport to cancel obligations owed to children, workers, or other beneficiaries. Their mutual release cannot necessarily extinguish independent claims held by those persons. Conversely, third parties who benefit from the relationship cannot presume that their disappointment authorizes them to dictate its continuation. The proper authority to waive or enforce a claim may not reside in the most emotionally invested observer.
This two-sided constraint does not explain every disputed outcome. It tells us which inferences are insufficient, not exactly what settlement is fair. There may be cases in which a particular actor really is obligated to provide substantial assistance after leaving, even where doing so is inconvenient. There may be cases in which no private person bears the full responsibility for preserving continuity, and the remaining demand is a matter of public justice. The constraint serves as a guardrail for the specification of duties rather than a complete normative formula.
4.15. Rival theories and the originality burden
A critical examiner could argue that nothing in this chapter requires a separate theory of temporal jurisdiction. Betzler already combines relational freedom with residual duties; Lopez-Cantero challenges the imposition of intrusive deliberative requirements; Holm defends enduring obligations across transformation; Hohfeldian analysis distinguishes liberties and powers; Shiffrin explains the moral significance of delegated duties; Räty distinguishes wronging from simple rights violation and investigates the victim's standing concerning amends. Care ethics supplies the wider social context of dependency, while moral residue theory already separates justified choices from the regret associated with their consequences. A combined account may explain every case we have examined.
This objection is not a peripheral concern. At the present stage, it is substantially correct. The chapter's work is primarily a critical reconstruction and a clarification of what an original contribution would have to establish. The phrase “normative life of an ending” names a research focus. It does not create a new philosophical category simply by naming it. Claims concerning the difference between intimate continuation and financial support, the persistence of duties after dissolution, and the appropriateness of regret without culpability have extensive precedents.
What might remain for the dissertation? One possibility is a rigorous account of what makes assertions of discharge valid in cases where relational change is permissible but several parties dispute which duties survived. Another is an account of the moral authority to determine transitional arrangements where the relevant claims are neither straightforwardly promissory nor reducible to general beneficence. A third is a focused theory of how institutions misidentify the historical provider of a benefit as its continuing guarantor, especially where the distribution of responsibility was never legitimately authorized. But the author's own The Moral Remainder and When the Gift Becomes Capacity already cover adjacent terrain. Each prospective contribution faces the further test of internal originality as well as external scholarly precedent.
The appropriate research strategy is therefore adversarial. For each candidate principle, construct a case in which the claims are genuinely incompatible, give the strongest rival theories their full resources, and identify whether any distinctive conclusion survives. A mere assertion that existing accounts are “incomplete” will not do. The dissertation must show how they fail to resolve a specified issue, why the failure matters, and whether the proposed remedy has fewer unacceptable consequences. Until that test succeeds, the work remains a philosophically serious investigation rather than a proven theoretical discovery.
There is no shame in that status. Doctoral scholarship often begins with a question whose best answer is not known. What would be indefensible is to fill the space with elaborate claims of originality before the argument has earned them. This chapter's contribution to the research process is to narrow the terrain: the problem is not whether permitted endings can leave obligations. They plainly can. The difficult problem is how to specify and legitimately alter those obligations without confusing the survival of a need with the survival of one person's authority to command its satisfaction.
4.16. The place of this chapter in the dissertation
The analysis of endings adds something that a theory of how duties arise cannot supply by itself: a way of distinguishing the end of one form of participation from the survival of another obligation. Adrian may end a romantic relationship while retaining a financial duty. Mara may stop providing unsafe care while still owing information needed for a reasonable handoff. In neither case does the recipient's continuing hardship prove an unrestricted right to the former relationship. The relevant duties must be examined by their particular grounds and objects.
That conclusion brings us to succession. If Adrian's financial commitment can survive without romantic participation, who may fulfill it on his behalf? If another caregiver replaces Mara, does the new provider acquire Mara's duties, undertake different ones, or simply perform some of the same tasks? Chapter V distinguishes the work being done from the authority and accountability under which it is done. The difference matters most when the visible service continues and everyone assumes that responsibility has moved with it.
The inquiry into succession should also be separated from a comprehensive social theory of care. The existence of public obligations is indispensable to the fairness of particular transitions, but the dissertation cannot resolve every question of distributive justice in a chapter about the ending of relationships. A focused account must be explicit about what it assumes concerning rights, institutional duties, and the legitimacy of role assignments. These background theories may be contested; their contestability does not license the manuscript to replace them with intuition.
Finally, the distinction between ethical and legal judgment must remain visible. Moral permission to leave does not automatically entail legal permission to terminate contractual or statutory responsibilities, and legal enforceability is not a complete measure of what people owe one another. Where a case depends upon professional regulation, family law, confidentiality duties, or disability services, the relevant rules require jurisdiction-specific verification. The philosophical argument may ask what those rules should be; it must not invent them.
4.17. Conclusion: ending is an act, not an erasure
An ending may therefore be legitimate and still consequential. Morgan can suffer a serious loss without possessing a right to Adrian's continued affection; Clara can still need care when Mara's health makes direct provision unsafe. In each case, the law or professional rules may add duties that moral reflection alone cannot specify. An adequate judgment must acknowledge the hardship without turning it into automatic proof of wrongdoing—and must remain alert to independently grounded responsibilities that a simple account of personal freedom would overlook.
Adrian's departure does not erase Morgan's sacrifices; Mara's illness does not make Clara's needs less urgent. Yet neither hardship determines, by itself, exactly what Adrian or Mara must do. We need to ask which earlier commitments still bind, what each requires, and whether a duty has been altered, released, or separately undertaken by a successor. The past does not own the future. Some of its justified claims nevertheless survive within it.
The central danger runs in both directions. An agent who has made a binding commitment cannot cancel it by declaring the relationship over. A beneficiary cannot transform every disappointment or dependency into authority to require continued personal participation. The right question is narrower and harder: what particular act remains owed, and who has the standing to demand it?
Chapter V examines the same problem from the other end of a handoff. A retiring clinician can send records, and a new practice can be capable of treating the patient, without the new practice having accepted ongoing care. That apparent continuity may conceal an unsettled question of duty. To examine it, we must distinguish delegation of a task, the creation of a new obligation, and the valid discharge of the old one.
Continue reading
← Previous chapter · Next chapter → · Full contents
This is a developmental scholarly manuscript, not peer-reviewed or defended. The bibliography and research-status statement are accessible from the main contents page.