Chapter Seventeen treated departure as a reality check: after the relationship ends, whether the materials can stand independently, whether unresolved disputes can continue to be verified, and whether Ye can complete the handover without accepting a single narrative. The exit is thus already open, yet it does not automatically restore the whole of judgment and the capacity for living. Ye still needed housing, income, assistance with the files, emotional containment, and professional opinion. Aid is not the dispensable business of a bystander; but supporters who control resources, language, and procedure may also, in goodwill, acquire total interpretive authority over Ye's experience. The service station assigned the advisor Su to assist Ye, Cen provided temporary lodging, and the peer team offered short-term work. All three wished to reduce risk, yet each raised different demands: Su suggested that all communication pass through her, Cen believed Ye should make the project's problems public, and the peers hoped to confirm as soon as possible that he would bring no controversy. Each demand may have a legitimate object, and each may also expand aid from limited support into new conditions of loyalty, transparency, and obedience.
This chapter does not oppose professional proxy work, emergency protection, or the advice of family and friends, nor does it make specific judgments about real medical, legal, and crisis cases. When exhausted, in danger, or short of information, people may indeed need others to act on their behalf, and denying this need would turn subjectivity into isolated self-rescue. The harsher possibility still deserves to be said plainly: helpers can use the language of safety, recovery, professionalism, or justice to monopolize information, prolong crisis identity, and interpret the refusal of advice as the problem itself. The critique must make aid verifiable, limited, and exit-able, rather than writing protection as a new scrutiny of the aid recipient.
Aid Tasks and Proxy Authority
Ye asked Su to verify the file procedures, authorizing only a single professional task. Su may ask for information relevant to the contract, the deadlines, and the security of the materials, but she cannot thereby acquire authority over all of Ye's friendships, emotions, and career decisions. The clearer the task's scope, the better both sides can judge which information is necessary, when the task is complete, and whether new problems require renewed authorization. At times a subject cannot state precisely what he needs; the helper may clarify it jointly rather than mechanically waiting for complete instructions. Clarification should still present options, give reasons, and permit refusal; professional judgment may not directly convert a local request for help into total management.
Su held that the best goal was a safe departure, while Ye wanted to preserve his responsibility to the interviewees and the possibility of future cooperation. Risk assessment may prioritize certain actions, but long-term goals cannot be set by the helper alone. The two sides can distinguish immediate protection, fact-checking, life stabilization, and relational choice, each with a different authorization and time frame. Where legal requirements or the safety of third parties are at stake, aid cannot simply defer to Ye's preferences. Restrictions should state their object, their scope of authority, and their review; one necessary boundary may not be used to infer that Ye globally lacks judgment.
Ye was uncertain whether Lin intended control, and uncertain whether to go public. If Su required the prior adoption of a definite narrative before providing service, she would use resources to manufacture an epistemic convergence. Aid can proceed around known losses, the preservation of materials, and options, leaving cause and intention unresolved. The helper may equally say that she does not know, and refer out matters beyond her competence. Confidence is not a substitute for professional responsibility; the statement of boundaries in fact reduces dependence.
When resources are limited, institutions need to rank risks; if only those who use the most extreme language and disclose the most private experience can obtain help, subjects will be forced to inflate their narratives. Eligibility should turn on observable need and consequence, with multiple forms of proof and appeal. Immediate protection may set a lower threshold; severe attribution of blame requires stronger evidence. False applications harm other recipients and warrant proportionate verification. What is verified is the facts of eligibility, not the applicant's acceptance of the institution's interpretation of his person.
Protection and Proxy Must Be Layered
Su explained the appeal deadlines, the possible outcomes, and the risks to the materials, letting Ye decide for himself. This kind of aid chiefly expands understanding and does not choose for him. Information is also subject to framing, and should therefore state its sources, its alternatives, and its uncertainties, rather than presenting one professional opinion as the only reality. Too much content can itself become a new burden. The helper may provide, in stages, the information relevant to the current decision, and let Ye return to the raw materials at any time, instead of using simplified summaries to acquire an unchallengeable standing.
Cen accompanied Ye to meetings; she could serve as witness, help with the record, and call pauses under pressure. The presence of an accompanying person changes the scene, and may protect Ye, but may also make the other side more defensive. Roles should be stated beforehand: who speaks, who keeps the materials, and who is entitled to accept a settlement. The companion may not, because she has invested time, make commitments on Ye's behalf, nor interpret his hesitation as a need to take over. She may remind him of agreed bottom lines when necessary; the decision still returns to the authorizing subject.
When Ye was hospitalized with a high fever, he could authorize Su to receive documents on his behalf and apply for extensions. The proxy addresses limits of time and body; it does not mean that Su may decide whether to settle, to go public, or to change careers. A list of powers, an expiry date, and a record make the proxy revocable, and also let third parties know who may decide what. In emergencies, when full authorization cannot be obtained, the applicable institutional rules may permit the minimum necessary action on another's behalf. After acting, one should inform, confirm, and review as soon as possible, returning the capacity to the subject, rather than using a later safe outcome to justify every excess.
Where there is imminent and severe danger, some institutions may impose limited protection despite the subject's objection. Real-world judgments must rest on legal and professional standards; this thought experiment does not set them on its own. Normatively, it requires at least a specific risk, the least restrictive means, review by distinct roles, a temporal endpoint, and a place for the protected person's expression. Compulsory protection may not become a shortcut to interpretive authority. Restricting one dangerous action does not mean that the institution may decide all of a person's relationships, narratives, and future; once the risk subsides, the powers should end item by item.
Professional Authority and Dependence on Goodwill
Su knew the file procedures well and could explain authority and evidence; she did not necessarily understand the ethics of the footage, Ye's relational history, and his career aims. The further a professional opinion exceeds training and remit, the more it should be downgraded to ordinary advice. A title may not travel across domains, nor does a recipient lose his epistemic standing in other areas because he sought help in one. When several professional conclusions conflict, a second opinion and openness about method are more reliable than choosing whoever sounds more confident. Repeated consultation also has costs; the intensity of review can track consequence, unknowns, and reversibility.
If Su called Ye's hesitation denial and his revisions vacillation, the description could expand from a state into an untrustworthy character. Certain clinical concepts have legitimate uses within qualified practice, and this chapter does not deny them; in ordinary aid, cancelling testimony by borrowing terminology requires independent grounds and limits of scope. Emotion, memory, and behavior can be observed; cause and diagnosis cannot be inferred from a single relationship. Support should first address visible need; formal assessment proceeds separately, by its own standards.
The application form offered only "victim," "party to the conflict," and "voluntary withdrawal," and Ye's mixed experience hardly fit it. To obtain services he chose the nearest category, and the institution later cited that checkbox as proof of the whole truth. Classification serves allocation and should not become a historical confession. Allowing annotations, corrections, and limits on use prevents administrative convenience from rewriting experience. Fully open classification would raise processing costs. One may keep a limited set of options while stating that it answers eligibility rather than total truth, and provide an entry point for exceptions.
Su had once helped another person sever all contact and recover, and she concluded that Ye should do likewise. Past experience supplies hypotheses; it does not answer for the current materials. Persons, relationships, duties to third parties, and risks differ, and one outcome cannot become the standard path of recovery. Aid institutions should record failures, exits, and cases of non-fit, lest they learn only from the successful who remained in service. Acknowledging the boundaries of a program does not weaken professionalism; it improves the capacity to match.
Goodwill Can Form New Structures of Dependence
Cen at once provided housing, transport, and entry to social life; when Ye refused her advice to go public, he feared losing all support at once. Cen may have made no threat, yet the structure still changed the cost of speech. Diversifying resources, making the rules of cohabitation explicit as unrelated to opinion, and offering alternatives reduce the chance that one relationship becomes a master switch. Gratitude is the value of a real relationship and is not an authorization. Ye should shoulder his shared responsibilities of living together, without repaying help by adopting Cen's interpretation.
The service station needed cases to prove the worth of its funding, and its staff also derived professional meaning from helping. These interests do not prove that the institution deliberately prolongs problems, but they should enter governance. Case use requires separate consent; service goals include the possibility of ending; and outcome evaluation does not count only the frequency of continued contact. Cutting institutional resources cannot be excused by appeals to conflict of interest either. Public funding, independent oversight, and the recipient's exit handle risk better than abolishing necessary services.
Ye joined a group of former members; the similarity of experience brought understanding. But if members who acknowledged that the project once had value fell under suspicion, the group would sustain belonging through a common enemy. Horizontal aid, too, erects its own gates of reality. Group rules should permit revision, minority positions, and departure, and not treat a more severe narrative as a credential. Joint advocacy needs a limited external stance; members may disagree without speaking for the organization. Coordination of action and unity of mind must be kept apart.
Cen had managed Ye's daily life for a long time; exhausted, she demanded control of his schedule and his contacts, perhaps to reduce her burden rather than from an intent to dominate. Caregivers need boundaries, rotation, and support; otherwise the system converts free stopgap labor into private power. Explaining the pressure does not remove responsibility for overstepping, but it helps locate the conditions that can change. Nor may Ye understand subjectivity as the unlimited calling on others. Both the one seeking help and the one giving it can refuse; public services reduce what private relationships must carry alone.
Materials, Outcomes, and the Recipient's Responsibility
Su could acknowledge Ye's fear, his damaged income, and the restrictions on the files while remaining agnostic about whether Lin had premeditated anything. If support must assent to a complete interpretation, aid will reward certainty. Layering what happened, what was experienced, cause, intention, and responsibility lets protection and inquiry proceed in parallel. The manner of questioning must weigh safety and relationship. Stabilizing first under imminent risk and reviewing complex attribution later does not lower the standard of fact; it lets different decisions use different thresholds.
Su's summary selected Ye's statements, questions, and institutional categories, and would become the starting point for later professionals. Records should distinguish verbatim speech, observation, inference, and decision, and allow Ye to see them, correct matters of fact, and attach dissent. A record is not reality itself because a professional produced it. Nor may protecting the record's privacy let the institution rewrite it in secret. Access, versioning, and downstream correction together maintain continuity.
When Ye found an email supporting Lin, the group said this showed that he was still under influence; any counterexample strengthened the same conclusion. Aid should allow the materials genuinely to change the scope, even if the overall risk judgment ultimately stands. Counter-evidence must be relevant and reliable, and one favorable item may not erase the other evidence. The stability of a relationship does not depend on conclusions never changing. The shared commitment is to revise according to the materials, not to guarantee each other that some story will remain unchallenged.
The service station may restrict the circulation of case materials, yet it should allow Ye to consult a lawyer, a doctor, peers, or persons of his own choosing. The object of confidentiality and the shareable scope must be clear. Requiring all external contact to pass through Su would let protection acquire an information monopoly. External supporters bear duties of privacy as well. An independent path does not mean publishing all materials; it means that necessary consultation is not subject to a single helper's permission.
Aid Outcomes Cannot Be Judged Solely by Compliance with Advice
Ye followed Su's advice and stopped contact, and the conflict declined for a time; this does not show that every exercise of information control was justified. The outcome may have issued from several causes, and the boundaries of right cannot be ratified retroactively by luck. Evaluation must consider risk, the subject's understanding, alternatives, authorization, and long-term capacity together.
If Ye chose to continue communicating and was damaged, the helper may say that she warned him, yet should still provide the emergency entry and the follow-up information required by her role. A service may not cancel basic help on the grounds that "you did not listen." Ye must also bear the choices within his control; responsibility is not absorbed entirely by one side.
Whether Ye can better examine materials, express what he does not know, manage boundaries, and use multiple sources of support comes closer to the aim of aid than the number of consultations. Long-term service is sometimes necessary, yet one should still ask at intervals whether the permissions can shrink, whether the gains are portable, and whether the relationship can end safely.
Evaluating by contact counts, retention, and case stories makes services likelier to prolong relationships; evaluating only by rapid case closure pushes toward premature exit. Metrics need to combine the actual tasks, the recipient's goals, the risk of recurrence, and independent feedback; success cannot be defined by a single number.
The Recipient Is Not a Party Without Responsibility
Ye may refuse advice, and he should also state workable arrangements when third parties and shared resources are involved. Subjectivity is not an exemption from reasons, above all when decisions affect the interviewees' privacy and colleagues' handovers. The helper helps form the reasons and does not bear all the consequences in his place.
If Ye exaggerated certainty to obtain help, he could misdirect the allocation of resources. He should distinguish what he remembers, what he infers, and what he heard, and correct the record when he finds errors. Exhaustion and fear affect expression; correction should be offered, and deception not concluded from a single discrepancy.
Harassment, leaks, or false accusations within a support group still require proportionate handling. Protecting recipients does not mean deleting the accused from the roster of subjects. Specific conduct undergoes verification; group membership is not cancelled wholesale for an individual error.
When Ye decided to end the service, he should retrieve his records, confirm the pending deadlines, and return resources; the institution should likewise explain the paths of return. Freedom of exit and a minimal handover coexist. Where the relationship raises immediate safety concerns, the manner of ending may be adjusted by professional institutions; this chapter does not offer real-world arrangements.
Complaints and Shrinkable Aid
If Ye believed that Su had overstepped, he could not be left with a complaint addressed to Su alone. Distinct personnel, anonymous alerts, and external oversight let materials bypass the accused. A complaint does not stand automatically; it still requires facts, response, and proportionate handling.
The service station had labeled Ye "refusing to cooperate" and later found the classification improper; deleting it internally was not enough to repair the harm. The housing and employment agencies that had received the records should be corrected, and Ye should know the scope. Where all copies cannot be recalled, the residual risk should be stated honestly.
Goodwill does not automatically cancel the losses of income, relationship, and reputation caused by unnecessary restrictions. The intensity of liability turns on the risk at the time, the materials, the powers granted, and the alternatives available. Acknowledgment and compensation do not punish the act of helping; they let high-authority aid bear proportionate consequences.
A single complaint may not publicly define Su. She has the right to see the charges, to explain the emergency context, and to receive a layered disposition. If the institution substitutes individual punishment for the repair of rules, staff will not dare to act under risk, and the original structure will remain.
Building Aid Relationships That Can Shrink
The service station set file consultation, communication proxy, life support, and public advocacy as separate authorizations, and Ye chose as needed. Expanded use requires renewed specification; emergency exceptions expire automatically. A list of powers cannot replace everyday relational judgment, but it provides a common object against which overreach can be identified.
Ye knew who funded the service, whether cases were used for publicity, and whether the advisors had dealings with the project. Interests do not prove bias, but concealment damages judgment. Disclosure revolves around the relevant powers and does not require helpers to surrender the whole of their private lives.
Su handled procedure, Cen provided life support, and the peers explained professional conditions; no one party represented the whole of reality. Disagreements were handled by Ye and by appropriate procedures, with a second opinion sought when necessary. Multiple centers add burden; a replaceable coordinator may consolidate them, but coordination does not own the final ruling.
Stating goals, reviews, and handovers at the outset lets the relationship end by task completion, by the subject's choice, or by a mismatch of capacities. Exit is not interpreted as betrayal, and the institution does not prove care through unlimited service. Long-term support may still exist, with a specific reason for each extension.
Returning from the Protected Party to a Subject in Relationships
Whether aid can shrink must also be tested when the relationship meets disagreement. That Ye decides for himself in calm periods does not prove that Su would still respect boundaries when he refuses advice, revises the narrative, or complains about the service. The genuine position of a subject shows itself thus: support is not withdrawn at once upon disagreement; necessary resources are not bound to loyalty; and the helper can acknowledge that her own judgment may in turn become an object of review.
Ye refused to let Su receive all his mail; he may have underestimated the risk, or he may have been protecting privacy. Su should state the concrete consequences, the alternatives, and the bounds of what she cannot in conscience do, rather than writing the refusal itself as a lack of judgment. If the risk does rise, she may narrow the service or initiate appropriate procedures, and the decision must still remain tethered to its object. Nor may the recipient demand unconditional execution from the helper. Where Su believes that an act is unlawful, unsafe, or beyond professional duty, she is entitled to refuse and to refer onward. Both refusals need a scope, and neither may expand into a total verdict on the person.
Su misread a deadline and corrected it promptly; Ye may pursue the losses caused without therefore treating all professional opinion as deceit. The scope, cause, consequence, and repair of an error are judged separately. Requiring helpers to be perfect in order to remain credible encourages concealment; demanding that Ye not pursue the matter out of goodwill cancels accountability. Institutions should make correction ordinary work rather than professional humiliation. Gross negligence and deliberate concealment carry different liabilities and must not be diluted in the language of learning.
Ye feared that complaining about Su would lead the service station to withhold further help, or damage the resources of others seeking help, and so he stayed silent. The institution should separate complaints, current services, and funding eligibility, and provide external entry points. A complainant need not publish the whole of his experience in order to point to a specific overreach, nor does complaining automatically confer the status of being factually right. Colleagues' concern for Su's work pressure may supply context, but it cannot answer the materials in her place. If organizational solidarity requires denying internal error first, the aid system becomes the very object it cannot help.
After Su acknowledged the overreach, corrected the records, and handed over, Ye might continue with limited consultation or end the relationship. The institution may not treat reconciliation as a metric of service success, and Ye may not use the refusal of reconciliation to demand unlimited punishment. Repair addresses rights, records, and consequences; whether intimacy and trust return belongs to the two parties of the relationship.
Multiple Subjects and Learning After Aid
Ye needed help, and he remained answerable to the interviewees, his colleagues, and his public statements. Writing him only as a passive victim shrinks his capacities; returning all responsibility to the individual deletes resources and the differentials of power. Institutions should record what he can decide, which matters temporarily require a proxy, and which consequences the organization bears, updating the record as capacities change.
Su's caseload, training, supervision, and rest affected her judgment. Demanding unlimited empathy and constant availability would mask labor conditions behind a moral mission and ultimately hand the risk of exhaustion to Ye. Reasonable workloads, team review, and clear referral protect the helper and the recipient alike. Labor conditions are not an excuse. Those who hold professional powers remain answerable for overreach and negligence; the organization answers for the resources and rules that produce systematic error.
Cen could accompany, house, and offer ordinary advice; she should not, by proximity to Ye, come to bear crisis assessment, custody of evidence, and round-the-clock surveillance. Pushing professional duties onto friends exhausts intimacy and leaves errors with no formal bearer. Public services should assume the high-risk duties, and private relationships keep their voluntary range. A friend who perceives imminent danger may still call for help, and may not withdraw entirely from concrete moral responsibility by pleading "I am not a professional." Responsibility is proportionate to capacity, to what is known, and to feasible action.
Lin, facing the complaint, has the right to know the specific charges, to answer the materials, and to be free of group shaming. Protecting Ye cannot be accomplished by cancelling Lin's factual and procedural standing. If risk requires temporary limits on contact, protection may come first and appropriate procedures judge afterward; the scope of restriction does not automatically prove final responsibility. Treating Lin precisely also protects Ye. Only when conclusions are verifiable and responsibility is layered does Ye's experience avoid being negated as a whole because one overstrong accusation fails.
Public Learning After Aid Ends
When several recipients met the same archival threshold, the service station could aggregate the mechanism, the decisions, and the consequences, and press for changes in the rules. Public learning need not display identifiable stories, and case use requires separate consent. Recipients are not raw material for institutional branding and policy advocacy. Learning nothing at all lets the same problems repeat; preserving too much builds an archive of the vulnerable. Minimal data, controlled access, time limits, and public reporting on mechanisms can hold both in view.
Whether Ye can obtain materials himself, find multiple entry points, state his unknowns, and refuse safely says more about the fruit of aid than whether he keeps contact with Su. Some long-term support does not aim at independent handling; still, one should ask whether the subject holds choices within the range of the possible, rather than shaming those who need continuing care with a standard of self-reliance. The transfer of capacity also includes knowing when to ask for help again. Depending on no one at all is not the standard of maturity; being able to choose, compare, and end particular dependencies comes closer to subjectivity.
Where the service station cannot provide housing, medical care, or litigation representation, it should state its scope and give credible referrals. Recording only completed services makes unmet needs disappear. Refusal may be reasonable, yet it should become material for resource planning and risk assessment. A referral is not pushing a person to another address. Important matters require confirmation of receipt, protection of the material's use, and a preserved path of return in case of failure.
To prove that aid works, institutions might require Ye to disclose relationships, emotions, and his subsequent life. Audits should give priority to task outcomes, anonymized sampling, and voluntary feedback, collecting only the minimum material relevant to decisions. A recipient's refusal of tracking should not lower future eligibility for services. High-consequence complaints require the preservation of evidence, which differs from pervasive performance surveillance. Only when uses are kept apart can accountability coexist with help without turning help into continuous observation. In the end, the service station narrowed Su's role to advisor on files and deadlines, and Ye decided for himself whether to communicate with Lin; the rules of Cen's lodging concerned shared living only and were not tied to public stances; the peer team evaluated by work materials and did not require Ye to guarantee a future free of controversy. The supporters kept the necessary records separately, and Ye could view, correct, and end the authorizations. The interviewees' privacy and the shared handover remained protected. Such an arrangement does not guarantee that Ye chooses correctly, nor can it erase relational debts, professional error, and scarce resources. It gives errors multiple positions of discovery, and turns the helper's capacities into limited services rather than total personal authority. The subject is not the wholly self-sufficient person, but the one who, within relationships, can still know who is helping and why, what is being refused, what is being borne, and who reclaims the decisions when capacity returns.
The gravest risk of aid does not come only from malice. Goodwill, professional success, the concentration of resources, and unending crisis suffice to expand a support relationship; recipients, too, may use a fragile identity to evade responsibility toward others. Only a critique that preserves need, power, and responsibility together will neither cancel aid in opposing control nor replicate control in providing it. The next chapter enters the third act and discusses how to begin again from local predictability: recovery does not require a fully stable world, but only that a few rules, relationships, and bodily rhythms again come to support action and revision.