Decisions under coercion: discussing choice and its constraints

Epigraph

“The apparently lesser evil.”

Samuel Pufendorf, On the Duty of Man and Citizen, Book I, Chapter I, §16, paragraph on mixed actions. A short fragment from Frank Gardner Moore’s English translation. It concerns what appeared preferable to the person in an emergency, not a finding that the decision was harmless. Primary context.

What this article is about

In wartime, “choice” can conceal different conditions. A person faces threats, demands to hand over a document, an answer linked to a relative's safety or several painful alternatives. Afterwards comes an accusation: you agreed yourself. Yet an outwardly completed act does not establish how available refusal was or who made it dangerous.

This article considers coercion, necessity and responsibility without deciding guilt from a distance. It distinguishes physical force, threats, absent means, ordinary inconvenience and a limited next step. Wartime scenes are hypothetical; no route, resistance tactic or legal regime is prescribed. Historical interlocutors help pose questions rather than replace competent contemporary assessment. No dangerous refusal, detailed disclosure or story of growth is required to prove freedom. Continuing danger gives protection and actual local help priority. Assistance for distress and legal advice remain different tasks, with their own conditions and purposes.

I. Describing an action does not yet explain its conditions

Begin with a hypothetical wartime scene without details of violence. Someone handed over a document. That description alone establishes neither personal initiative, an understood agreement, fulfilment of a role nor response to a threat. A hand's movement reveals no attitude towards every alternative. Consideration concerns both the event and how acting otherwise was made possible or unavailable. Document contents and participants' names need not be disclosed immediately to an arbitrary listener. Suitable conditions and a competent channel may matter. Action and its conditions belong to the question of responsibility itself, rather than an optional qualification after a verdict already delivered. A limited account can preserve that question without pretending to contain every necessary fact.

Conditions should not be guessed from facial expression or the confidence of an account, however. A calm voice does not itself prove willingness, and visible agitation does not determine its cause. A more precise description can be retained: the transfer happened, while information about available paths remains incomplete. This neither cancels the event nor declares every explanation true in advance. It leaves room for significant facts without which a general conclusion about freedom would be premature. Sometimes those facts remain insufficient even after a conversation; uncertainty then remains part of an honest conclusion.

It also matters to distinguish an action from endorsement. Someone may carry out what they do not consider good if they see only an even more difficult alternative. Another person may support the intention without taking part in execution. Neither should automatically be treated as the other. In the hypothetical example, handing over the document does not necessarily mean sharing every purpose of its recipient. Yet a lack of endorsement does not automatically establish that there was no choice at all. Sequential reasoning begins here: establish the event, learn the significant conditions, distinguish participation from attitude, and acknowledge what is unknown. Only then can a particular contribution be discussed rather than a general label for the person. This sequence does not promise a convenient answer; it protects the question from replacing an observed act with presumed complete consent.

II. Physical force differs from deciding under pressure

If someone directly uses another person’s body as an instrument of action, that differs from a situation in which the person still makes a decision, even under severe pressure. Imagine two hypothetical scenes without details of violence. In the first, a stronger person physically moves another’s hand. In the second, they threaten a grave consequence while demanding a particular act. The outward result may be the same, but the mechanism differs. In the first description, bodily participation does not establish ownership of the intention; in the second, the threat and the actual availability of refusal need consideration. Coercion remains significant in both, and the distinction does not make a threat insignificant.

This distinction requires no examination in visible resistance. An absence of resistance known to an observer is not, by itself, sufficient evidence of safe, freely given consent. That is a limit to inference rather than a diagnosis of the harmed person’s condition. Historical ideas about which outward signs must accompany coercion cannot be imposed on every person as a requirement to prove their experience. One scene does not disclose all the conditions. Requiring a dangerous refusal to make an account convincing would add a new cost precisely to the person whose space is already constrained.

The distinction also matters for the responsibility of whoever creates pressure. The person who formed the dangerous situation does not disappear from its description after another’s act. Yet we need not establish every legal consequence in advance or call every strict request a threat. That is a different level of work with specific facts and applicable rules. Here it is enough to preserve the philosophical subject: an outwardly performed act does not prove that the person freely chose all its conditions. When circumstances require separate examination, a general moral slogan should not replace it. Precision begins by acknowledging that forced bodily control, responding to a threat, and ordinary fulfilment of a responsibility call for different questions. Respecting that distinction does not require the harmed person immediately to explain or assess themselves.

III. A threat changes the cost of refusal

A threat matters as more than an unpleasant phrase: it changes the anticipated cost of an answer. In a hypothetical wartime setting, information is demanded and refusal linked to a grave consequence for someone close. Outward agreement cannot be described like an answer to a genuinely voluntary request. Who could cause the consequence, what was known and which possibilities remained need consideration. This article does not assess a particular threat and does not propose testing it through action. Risks differ, but cannot be treated as absent merely because someone nevertheless said yes. Their answer does not turn a danger created by another into an unconstrained personal wish. Conditions must remain part of the account rather than disappearing behind agreement.

Discussion requires specificity. What exactly was made dependent on the answer? Who could cause the consequence? Was another path accessible, how much time remained, and what information was known? These questions are listed here as boundaries of reasoning rather than a mandatory procedure for questioning a harmed person. In a small, safe situation, someone may choose one significant question. Under continuing danger, reflection does not require arguing with the person making threats or experimenting with their readiness to carry them out. The possibility of analysis does not imply a duty immediately to gather proof of one’s position.

At the same time, a grave cost does not let us mechanically describe every subsequent action with one word. Sometimes a threat concerns only a particular response while a limited possibility remains to choose the time, means of transfer, or an available intermediary. Sometimes no such possibility exists. Preserving the distinction means neither demanding complete freedom where it was absent nor assuming a complete absence of every decision without knowing the situation. This is not a symmetrical contest about who is more to blame. It is an attempt to retain the actual extent of action. An available small fork does not justify a threat created by someone else; the threat itself does not disclose precisely which forks existed. The constraint needs describing first, before discussing what remained possible.

IV. Necessity does not always have an author

A constraint can arise without a particular person deliberately coercing another. Suppose communication unexpectedly stops during war and an agreed task cannot be completed in its earlier form. Postponement, a revised agreement or partial participation may be considered. None is declared safe here; actual conditions need checking. Yet an absent means alone does not establish that the other participant deliberately removed it as pressure. Necessity arising from circumstances differs from a threat. This distinction can remain without blaming an unknown culprit by default or calling everything inevitable because the convenient initial route disappeared. It preserves a factual question about how the constraint arose, rather than resolving it through the most emotionally convincing account.

Someone may retain a decision among unfavourable alternatives. They choose not the desired option in general but what they consider permissible now. Such a choice cannot be called free of constraints; neither can it be described as the person automatically disappearing from the action. They can discuss consequences, consider others, and prefer to stop. Yet their response should be assessed through the situation in which it arose, rather than an imagined scene with a complete set of resources. If the original conditions are no longer available, remaining faithful to them does not restore communication or create the missing resource.

Necessity should not become a universal name for every difficulty, either. Sometimes a circumstance really cannot be changed in the required time; sometimes it is declared unchangeable without examination; sometimes an available solution requires effort no one has yet been willing to accept. Only a specific examination can distinguish these. An organizer may legitimately have limits, but those need naming: a deadline, a missing resource, or a commitment to third parties. A statement of inevitability without content closes the question before consideration. A more precise acknowledgement shows what is accepted as fact, what is a preference, and what remains unknown. It does not guarantee another option, but it prevents an inadequately explained arrangement from becoming a supposed natural law. A person retains the possibility of thinking about the particular situation even when the desired outcome is unavailable.

V. The historical source helps pose a limited question

In Book I, Chapter I, Pufendorf connects discussing responsibility with what lay within a person’s capacity to act. In §16 he distinguishes direct compulsion from choosing an unpleasant path under necessity, while §23 considers the limit of capacity with a qualification concerning impossibility previously created by the person. This is a limited reading of seventeenth-century natural-law philosophy. It does not establish contemporary legal rules, investigate responses to danger, or verify our hypothetical scenes.

We value a question that can be placed beside this argument: which particular action really was available, and who formed the condition? The answer must remain specific. One general idea cannot establish a duty to overcome any threat through strength of character. The historical text has its own religious foundations and views of human qualities. They do not become contemporary psychology because a philosophical question remains significant. In particular, a description of outward resistance must not become a universal duty imposed on a harmed person or proof that consent was free when such a sign was absent.

The qualification concerning previously created impossibility also calls for careful independent consideration. In a hypothetical project, someone could have promised to supply material and then made it inaccessible themselves; lost access for an unknown reason; or never possessed it. These are different descriptions rather than three degrees of one character trait. Present impossibility does not erase the promise’s history, but a past promise does not create a resource now. An earlier action can be discussed separately from what can no longer be done. This question preserves responsibility without demanding the impossible. It also does not allow an observer to invent earlier carelessness to avoid considering the present constraint. The source serves as a historical interlocutor; the boundaries of contemporary examples and conclusions remain the authors’. We do not declare the text a ready solution to every difficult case.

Such limited reading leaves room to disagree with the source. Respecting a historical interlocutor does not mean repeating every assessment. We can clearly name the question being used and separately state a contemporary authorial boundary. A dispute about attribution, for example, must not become a demand that a harmed person immediately recount everything, while their silence becomes proof that there was no threat. Someone may not wish to discuss the event with this listener. The listener then lacks sufficient grounds for a complete conclusion. Acknowledging this limit preserves both the seriousness of the event and the other person’s right to decide their participation in a conversation. Reasoning remains open to facts but does not appropriate them where they have not been provided.

VI. Choosing a less difficult path does not make it good

Under constraints, someone may prefer what seems less difficult. That means neither harm disappeared, complete approval was given nor the outcome proved best among every possibility. Assessment occurred with particular information and time. In a hypothetical communication failure, a brief available message might be preferred to an impossible promise of immediate completion. Safety and content need their own conditions; no universal format is recommended. Some questions remain open. A chosen path may try to avoid greater difficulty without proving that everyone became comfortable. Choosing among painful options does not make each a desired goal beyond the situation. Preference at that moment need not become approval of the condition imposed.

It therefore helps to distinguish preference under those conditions from a personal goal outside them. Someone may agree to a temporary arrangement they do not wish to retain later. If a resource subsequently appears, a new discussion does not necessarily express inconsistency. The situation to which the decision applied has changed. A temporary concession does not become indefinite permission to impose the same cost. The participant may nevertheless have specific commitments, and reconsidering them affects others. Consent for a limited period should be described accordingly, so that complete loyalty is not later demanded through endless repetition of an inconvenient action.

This distinction does not remove the question of consequences for third parties. One person’s unfavourable choice may affect another who took no part in creating the difficulty. Their experience should not be declared insignificant merely because the decision seemed compelled. Harm can be acknowledged while preserving the question of the acting person’s freedom. These are two subjects capable of existing together. Acknowledging harm is not a general verdict on someone’s whole person; limited freedom is not an absence of all concern for consequences. Sometimes only a small later contribution is available, sometimes a separate decision is needed, and sometimes information is insufficient. We obtain no ready formula. We do, however, retain the possibility of discussing what happened and available conditions without turning one person into the final explanation for the whole event.

VII. Ordinary inconvenience should not automatically be called coercion

Preserving the significance of a threat requires distinguishing it from the ordinary cost of an agreed action. Work can take time, a conversation can require attention, and an accepted responsibility can limit the next choice. Unpleasant effort alone does not prove that a decision was taken away. Imagine a hypothetical agreement: a participant earlier consented to prepare a short text by a particular date, understood its scope, and had access to material. Later they wish to do something else. The wish matters but does not automatically rewrite the original conditions. The question now concerns reconsidering a promise rather than necessarily being coerced when it was accepted.

The word ordinary nevertheless must not conceal an unknown grave cost. For one person an action fits available resources; for another it conflicts with an important commitment. The extent of constraint cannot be determined from our own schedule. We need precisely the conditions that change a response’s availability. Personal inconvenience can be a reason to decline a new offer even if it is not called coercion. Someone need not first prove danger to have the right to decline a voluntary request. Distinguishing kinds of constraint preserves that right too: refusal is possible without turning every unwillingness into an account of a threat.

The extent of inference matters. If an agreement had understandable conditions, that does not prove unconditional freedom in every other respect. If pressure was discovered, it does not mean every involvement in the same undertaking now has an identical character. A particular request, an earlier commitment, and a new demand can be separated. That description is more useful than a general argument about whether someone was free at all. It allows the specific cost of reconsideration to be discussed, boundaries to be maintained, and what had already been accepted to be acknowledged. Responsibility remains connected to actions over time rather than an unchanging label. We do not promise an easy conversation, but we reduce the chance of distinct events being quietly combined under one convenient word.

VIII. Responsibility for creating a constraint remains a separate question

Where someone creates a grave cost of another's refusal, their role does not dissolve into the final response. Imagine an organisation during war where a manager demands action through a threat and later describes compliance as entirely free agreement. Small remaining decisions do not remove the threat-maker's part. Personal contribution and how the demand was arranged remain questions together. The article determines no order's legal status or particular organisation's position. It prevents only the last person in a chain from being noticed while decisions making their answer dangerous disappear. Invoking duty replaces no such examination. Different participants' powers and knowledge still need their own consideration rather than one verdict for everyone.

Different people’s participation should not be turned into identical responsibility by default, however. One set the rule, another transmitted information, a third performed an available part, and a fourth did not know about the change. Each question requires its own information about action and possibility. Membership in a shared organization does not itself explain a role. Likewise, lacking complete authority does not establish a complete absence of contribution. In a small, safe situation, particulars can be named: what was assigned, who changed it, who knew, and which part remained available. This is an authorial aid to clarity, not a universal investigation protocol.

Precise separation of questions also protects those who want to help. A helper need not promise to remove a constraint created by someone else; the participant need not accept every proposed exit. Sometimes help concerns information, sometimes a change to the procedure, and sometimes merely not adding another demand. Such limited help cannot be considered proof of indifference, yet neither can it be declared a sufficient solution without examination. The promise’s boundary matters: which contribution is real now, and which conditions depend on someone else? This reciprocity does not cancel differences in authority and the cost of refusal. It allows discussion of one’s role without appropriating another’s decision or assigning the harmed person sole responsibility to repair the entire arrangement that constrained their action.

IX. Later knowledge should not quietly replace earlier knowledge

Afterwards, a means absent from earlier available information may emerge: another contact, possible assistance or changed communication conditions. This can help a present decision without establishing earlier availability. A possibility's existence generally differs from its use at the necessary moment. Suppose an important contact became known only after a decision under threat. A future plan can improve without the earlier choice becoming senseless merely through new information. Whether someone could have learned sooner remains a separate question. Present clarity should not silently become a complete set of earlier means and knowledge. A useful new possibility has its own time rather than rewriting everything before it.

At the same time, earlier ignorance does not justify declining every reasonable clarification in advance. Sometimes a person really had an accessible question that could change the decision; sometimes there was neither time nor a safe way to ask. These again are different circumstances. They cannot be assessed through one statement that someone could always have found out. What is established can be noted: which information was supplied, what was missing, and which inquiry remained possible. Where information is incomplete, the conclusion must retain its limitations too. What is unknown does not become proven carelessness because a simple cause is desired.

The distinction also matters in relating to oneself. Later clarity may help prepare another response, but it does not require replaying a dangerous episode as though today’s map of possibilities had existed then. Examining the past may be postponed or not undertaken alone. In a small, safe case, one transition is enough: a condition is now known, and it changes the next promise. Someone may acknowledge an error, a constraint, or remaining uncertainty without a general conclusion about their worth. Responsibility in this description concerns particular acts rather than endlessly examining whether someone was strong enough. New knowledge becomes a resource for a future decision; it need not become an instrument for condemning the person who acted under different conditions.

X. A next step is not a final verdict

We have distinguished the event, the form of pressure, existing alternatives, their cost, and when information appeared. A limited question about further action can now be asked. What really lies within an available step: communicate a change, decline a new demand, clarify a condition, acknowledge particular harm, or discuss nothing for now? The list does not exhaust possibilities or require an immediate choice. It separates preparing a next action from a duty first to settle the whole history of moral responsibility. Sometimes detailed examination is needed separately, but it should not artificially close an already available small possibility.

No next step proves that the earlier constraint has disappeared. Someone may explain the situation and receive a refusal; acknowledge consequences while lacking a means to restore everything; or state a boundary and encounter new pressure. The practical question therefore contains no promise of safety or successful agreement. Under danger, we do not propose testing someone else’s reaction to resistance. In an ordinary small agreement, one available action can be considered with its consequences and one’s right to stop. The possibility of acting within limits retains significance even without a guaranteed desired outcome.

Ultimately, responsibility becomes more precise when it requires neither complete power over circumstances nor the person’s complete disappearance behind them. Someone may be affected by coercion and still have a separate small choice; may themselves create a constraint and answer for that part of the history; or may encounter necessity without an intentional author. We do not reduce these situations to one assessment. Distinguishing them permits acknowledgement of harm, preservation of a threat’s significance, refusal to demand the impossible, and discussion of a particular contribution. Rejecting a general verdict does not cancel responsibility. It returns responsibility to actions, time, information, and available conditions. There we can seek the extent of a personal next step without requiring someone to prove their right to respect through a dangerous or inaccessible act.

What this means in practice

Begin with present danger. A continuing threat calls for available protection and actual local help, rather than a test of freedom through argument, experimental refusal or return to danger. This article prescribes no resistance route and replaces no competent assessment of conditions.

Without urgency, one small present question can be chosen without reconstructing a painful scene. Name an action and one condition changing availability: physical force, a specific threat, absent means or an agreed task's ordinary cost. Incomplete information can remain incomplete. Neither repeating a threat to prove it nor supplying a detailed history to an arbitrary listener is required.

Separately identify the task now: help for distress, factual clarification, legal advice or reconsidering a present agreement. Recipients and conditions differ. Before substantial details, ask about role, competence, information rules and confidentiality exceptions. NIMH describes finding help generally; actual local access and conditions are not promised. Receiving help determines no guilt and requires no completed moral dispute.

Where a small next step is available, name its actual result: communicate a change, clarify a condition or decline a new voluntary request. Agreement or safety is not guaranteed. Under threat, another person's reaction need not be tested experimentally. An earlier promise and third parties' rights still matter; a constraint cancels neither automatically.

Reflection can stop, especially if distress intensifies. Everything need not be written or told to earn support. A remaining question orders no immediate increase of risk or self-punishment. Present action and assessment of the past can require different time and competence. Their distinction preserves both seriousness and the possibility of appropriate help today.

Key quotation

“Things which exceed his powers.”

Pufendorf, On the Duty of Man and Citizen, Book I, Chapter I, §23, first sentence. A short fragment from Moore’s translation. The following qualification concerns capacity previously lost through one’s own fault; it does not permit inventing such a history where circumstances are unknown. Complete paragraph.

Connections between traditions

Sources	Question	Limit
Aristotle, Nicomachean Ethics III.1–2	How do action, pressure and a particular moment relate?	Ancient demands for endurance unto death are not transferred to someone harmed
Pufendorf I.I.16,23	Who created a constraint, and what remained available?	Historical natural law determines no present guilt; §23's qualification remains
Aristotle and Pufendorf	How can participation be discussed without erasing coercion?	Their systems are neither identical nor a substitute for a country's law
NIMH and present help	How can distress be addressed while past questions remain open?	An American resource guarantees no local services or confidentiality

Sources

1. Samuel Pufendorf. On the Duty of Man and Citizen, I.I, Frank Gardner Moore's English version. Sections 1–27 read in this task; title page checked: 1964 reprint, 1682 Latin basis. Two short quotations from §§16 and23. Historical philosophy is neither law nor psychology; the qualification about previously created incapacity remains.
2. Aristotle. Nicomachean Ethics III.1–2, translated by W. D. Ross. Both selected chapters read directly on 18 October 2026. This is new reading of a different English version; the library source's earlier Chase edition is not copied. Endurance unto death and historical judgments are not prescribed to the reader. No direct quotations added.
3. NIMH. Help for Mental Illnesses, reviewed April 2026. Urgent help, provider selection and list limits read. US contacts and local-access guarantees are not transferred.

Tags

war, coercion, necessity, responsibility, action, choice, threat, constraint, freedom to respond, Pufendorf, natural law, Aristotle, cost of refusal, physical force, available means, lesser evil, information, time, possibility, impossibility, promise, reconsideration, participation, role, consequences, safety, support, legal advice, NIMH, dignity

About the authors

Andrey Zarinsh is the author and founder of United Wisdom. He explores inner transformation, unity across traditions, intercultural dialogue and contemporary spiritual discipline. Instagram: @andrey.zarinsh.

Maria Evilkut is a co-author of United Wisdom. Alongside her research, she seeks to test these ideas in practice and observe how principles of unity and mutual understanding work in real life between people and cultures. Instagram: @ilkutki.