Responsibility for unintended harm

Epigraph

there are many differences in the particular cases.

Aristotle, Nicomachean Ethics, III.1, the concluding fragment of the discussion of particular circumstances in mixed actions. W. D. Ross's translation.

What this article is about

We can cause losses without wanting them. Helping some people changes others' plans; a useful event leaves someone with wasted expenses; a reasonable decision is carried out without considering those nearby. Afterwards it is easy to say: I meant well, so what happened has nothing to do with me. Yet intention answers only part of the question. It does not tell us what was known beforehand, which options existed, who was warned or what someone did on learning of consequences.

The opposite extreme requires responsibility for everything following our action, regardless of knowledge, influence or others' participation. Such responsibility has no boundaries and does not explain a specific case. The article distinguishes purpose, chosen means, foreseeable consequences, avoidable inattention and discovered losses. A hypothetical book exchange shows why lacking malice does not end discussion and a serious result does not automatically establish initial fault. Aristotle and Thomas Aquinas provide different grounds for examining voluntariness, circumstances, intention and due care. Their historical arguments supply no modern legal rules. The practice is an editorial exercise without established efficacy; it promises no elimination of every risk.

I. A good result does not absorb neighbouring losses

Consider an original hypothetical editorial story. Tamara organises a free book exchange in a local association of adult readers. The meeting is set for Saturday, but the venue unexpectedly becomes unavailable. It can be held on Friday evening, when the room is free. Tamara updates the announcement in the main group and is pleased: the exchange takes place and many people find books they need. Later, she discovers that some participants used a paper poster with the former date. Mikhail travelled from another district on Saturday, spent money on transport and missed another meeting. Tamara intended none of these losses. Her good purpose and his wasted trip nonetheless exist together.

The first task is to avoid choosing either part as the only truth. The exchange may have been useful while notification of its change was inadequate. Mikhail may really have lost time without Tamara having known about his journey or having had an easy way to warn him personally. An account of the event comes first: which announcements circulated, what changed and when, which channels were available and who was responsible for notification? Moral assessment should not outrun these questions. If we begin with defending the organiser or condemning her completely, facts readily become material for a predetermined conclusion.

Calling the loss a side effect describes its relationship to the purpose, not how little it matters to the person affected. Tamara's principal task is to preserve the exchange; Mikhail's wasted journey may matter substantially to him. Scale is not established by a consequence's place in somebody else's plan. Equally, calling losses significant does not equate them with every possible grave harm. Keep this case's own measure: travel, time, a missed meeting and possibilities for restoration. Such precision protects both participant and organiser from dramatisation that prevents discussion of actual actions. Majority benefit does not erase an individual subject, but that subject need not cancel the entire event.

Lack of intention must not be substituted for lack of participation either. Tamara decided, changed the date and chose how to communicate. These remain her actions, even though Mikhail's eventual journey also depended on his plans and available information. Participation in a causal chain does not establish equal fault in every case. It indicates where to look for one's own share of responsibility. Perhaps communication was reasonably complete; perhaps a list of poster recipients existed but was never checked. These possibilities differ. Saying that nobody wanted the inconvenience cannot settle them. An unwanted consequence can still be connected to the chosen method.

II. Intention, foresight and means

Intending a result and foreseeing its possibility are different relationships to an action. Tamara wants to hold the exchange. She may understand that some people will learn of the change late without wanting their wasted journeys. She may never consider the paper posters, or know about them but decide that updating the main group suffices. These distinctions matter. They require no guesses about hidden motives: available messages, actions and explanations are examined. If the extent of knowledge remains unestablished, say so. A confident moral formula supplies no missing evidence about what a person actually noticed and considered.

A harder distinction concerns harm called incidental although it is actually used as part of the means. Suppose a different hypothetical version requires expelling a reading circle already meeting so that the room becomes available for the exchange. Loss of access is then more than a neighbouring effect: clearing the room by cancelling another meeting enters the way the goal is achieved. Participants' additional expenses may remain unintended. The cancellation itself, however, cannot entirely be presented as an accidental inconvenience. In our principal scene Friday's room is free, so this substitution does not occur. The comparison shows why the path to a result needs describing alongside the final purpose.

A working question is whether we would obtain the good by the same method if this consequence did not occur. When the exchange uses a free room, Mikhail's wasted trip is unnecessary for its success. When success requires cancelling someone else's session, cancellation belongs to the plan. But this single question is no universal philosophical test and settles no complex case completely. It helps expose a possible substitution in an ordinary situation. Questions remain about the action's acceptability, the consequence's seriousness, available alternatives and obligations to affected people. The label side effect grants no moral permission and replaces none of these questions.

Foresight is not the affected person's consent either. Tamara might confidently anticipate someone arriving on the old date without having grounds to decide for them that their loss is acceptable. Where prior discussion is possible, their information may change the plan: someone left a telephone number, someone can contact their own circle, someone needs the new date early. We promise no universal agreement or preservation of the event's original scale. Yet participation has value of its own. Its absence does not always invalidate a decision; an available, convenient opportunity ignored needs explanation. Knowledge of a risk alone does not show how it was handled.

III. What was known when acting

In Nicomachean Ethics III.1, Aristotle examines voluntariness, compulsion, mixed actions and ignorance of particular circumstances. His distinctions include feelings and repentance; they are not reduced to a modern risk scale. The limited question for us concerns knowledge of a particular action, its object and conditions. Ignorance is not all alike, and an attractive purpose does not automatically make an action compelled. The ancient discussion supports attention to circumstances; the article applies that attention to its hypothetical rescheduling. Aristotle is assigned no ready notification procedure or modern assessment of legal fault.

Knowledge's timing matters. Information emerging after Mikhail's journey cannot automatically be placed in Tamara's mind before rescheduling. Hindsight readily makes the route, poster and missed meeting an obvious chain. When deciding, she may have known only about the room and main group. Reconstruct the available picture: did she know posters were circulating, were their recipients recorded, was there time for another message? These questions do not excuse inattention in advance. They establish precisely where it can be identified, rather than blaming someone for failing to know future facts that were not available to anyone yet.

But “I did not know” can conceal another question: should something easily discoverable have been checked? Tamara printed the posters herself and asked participants to distribute them. She could have remembered this channel without knowing every recipient's name. Having no list then means more than one action remains possible: asking distributors to notify their readers or leaving a clear message at the former venue. Such possibilities are considered alongside available time and energy. An impossible action cannot be demanded. But everything without a ready address book cannot be declared impossible either. Due care often begins with examining one's own ways of distributing information.

Keep three separate records: established knowledge, reasonably available checks and new knowledge after the event. The first helps assess choice, the second attention to conditions and the third the subsequent response to loss. They may yield different conclusions. Tamara might make no significant error beforehand yet respond poorly afterwards. She might notify carelessly and then address consequences responsibly. Later conduct does not rewrite the past, although it belongs to present responsibility. Separate assessment gives a more precise subject than labelling someone a good or bad organiser overall. It allows acknowledgement of a particular mistake without making it the person's final description.

IV. Good intention does not remove proportion

In Summa Theologiae II–II, question 64, article 7, Thomas Aquinas distinguishes an intended effect from another while retaining proportion's requirement. Article 8 examines chance and due care. These articles' historical subject is causing death, self-defence, sin and related distinctions, rather than organising meetings. We take limited questions about intention, proportion and care without converting the text into permission for modern legal cases. Its theological foundation remains its own, no mere copy of Aristotle's description of voluntariness. Nor does it yield universal permission for every action serving a good purpose.

For ordinary rescheduling, proportion means discussing the exchange's value alongside the interests affected. A few hours' inconvenience that can be announced beforehand differs from substantial commitments unknown to the organiser. No honest automatic conversion turns all these circumstances into one score. Available facts can be established: who was warned, which plans depend on the date, how necessary the new meeting is and whether another method could preserve its purpose. Assessment need not end in complete certainty. But it must concern actual interests, rather than nameless costs presumed irrelevant because they belong to someone else.

Examine alternatives without a false choice between one action and abandoning goodness altogether. Perhaps the exchange can happen later; perhaps some books can be passed on Friday while the main meeting is rearranged; perhaps the former date requires a notice at the venue even though the exchange will not occur. These options also have losses: books reach people later, volunteers spend additional hours, the venue is unavailable on a convenient day. Considering options does not mean that a less risky route always exists or is always preferable. It prevents calling the chosen means necessary before available methods have been compared. Necessity needs a concrete account of constraints.

Proportion concerns more than how many people suffer losses. An organiser may consider a small group an acceptable price for a larger group's benefit. But if that same group repeatedly bears expenses while others benefit, distribution becomes a question. Who decides whose time may be lost? Are people less visible in the main chat considered? Can participants state their dependence on the date beforehand? These questions give nobody an absolute right to stop shared work. They prevent a category of participants becoming the habitual place where costs are stored. Our editorial discussion considers who bears the burden, not merely the total number of successful exchanges.

V. Care before deciding and conditions for revision

Preventing unintended losses requires no initial knowledge of everything. Tamara cannot check every personal plan involving Saturday. But she can assemble a limited picture of dependencies: announcement channels, former date, access to the venue, distributors' contacts and the change's timing. Such examination belongs to her specific role. A casual visitor and an organiser have different possibilities for notifying others. Yet the role makes Tamara neither owner of others' lives nor guarantor of every event. Responsibility is related to action, available information, accepted promises and actual influence. The boundary must be described, rather than merely called reasonable.

Warnings received before acting have a particular place. If Mikhail wrote beforehand that he used only paper announcements and would arrive on Saturday, Tamara acquired specific information. If it was lost in a channel she could not continually monitor, that is a different case. If she read it but dismissed one person's loss as insignificant, that is a third. Their distinction is not established by the subsequent journey's cost: the result may be identical. Examine the warning's availability and the response. A real warning cannot be replaced by claiming organisers must sense every risk; received information cannot be dismissed as someone else's anxiety either.

Define conditions for revision beforehand. Rescheduling might be announced as provisional until principal channels are checked, with the final date confirmed after poster distributors answer. Or, if a certain group cannot be notified, the event might be rearranged afresh rather than brought forward. These decisions depend on the case and supply no universal norms. Their value is that new information can change the plan. If every objection after the first decision is called interference, risk examination becomes decorative. Information can be collected while being allowed to alter nothing. The problem then is no lack of information but no genuine possibility of revision.

Time constraints also need honest description. If a decision is needed within an hour, some actions are truly unavailable. But the hour cannot excuse omitting a simple message taking minutes or artificially ruling out later notification. Urgency has a specific subject: by when must the room question be settled, and by when can participants be warned? Those deadlines may differ. Deciding to reschedule does not end notification work. After choosing, a period remains in which consequences can still be reduced. Responsibility does not end at the moment the purpose receives an approved date.

VI. What changes when losses are discovered

When Mikhail reports his wasted journey, a new moral subject emerges. Tamara now knows a concrete consequence, not merely a risk. Even if her original action was reasonable, the report needs a substantive response. This does not automatically require accepting complete fault and paying everything the participant names. But Mikhail cannot be required first to recognise the event's benefit before being allowed to describe his losses. A good result for others is no condition of his right to give an accurate account. The conversation begins with the journey, announcement used and possibilities for restoration, while disputed details remain open to clarification.

Distinguish acknowledging a loss from finally deciding its distribution. Tamara can say: you arrived on the former date, spent time and money travelling, and I understand that the group update did not reach you. This does not establish which expenses relate to organiser error and which to circumstances or others' actions. It avoids making Mikhail prove his right to be affected at all. Accepted arrangements, information's availability and possible compensation can then be discussed. This is moral and practical examination of an ordinary situation, no legal determination of damages. Any external rules have their own subject where applicable.

Sometimes repair is readily available: passing on books set aside, helping find another meeting or reimbursing a specific expense through a shared agreement. Sometimes lost time cannot be returned. A request for restoration must therefore not include a promise to erase the event completely. Its irreversible part can be recognised while remaining consequences are reduced. People still liable to arrive on the old date also need notification. If Tamara concentrates only on disputing with Mikhail, she may miss another available action. The discovered case gives grounds to check whether the same way of causing loss continues, rather than demanding proof of absolute safety for all future events.

Further actions can make an intention more credible without proving previous faultlessness. Tamara may genuinely want to help and still omit necessary notification. Mikhail may recognise her care while retaining an objection to rescheduling's method. The parties need not exchange complete justification for complete forgiveness. It matters more to establish what can be done now and what will change next time. If disagreement remains, describe it: which facts are agreed, which obligations disputed and which measures proposed? Such a record retains an open subject better than requiring the conversation to end in general acknowledgement that nobody was at fault.

VII. Results do not establish everything about choices

Two identical checks may end with different consequences. Tamara might notify every known channel, while an old poster survives with someone nobody knew about. Elsewhere she might omit an obvious channel, but by chance nobody arrives. The absence of loss in the second case does not prove due care; loss in the first does not prove its absence. Results still matter for helping those affected and making later changes, but the original decision requires evidence of the available picture beforehand. We do not make actions independent of all consequences. We distinguish consequences as a subject for restoration from knowledge as a subject for assessing choice.

This prevents two convenient conclusions. Organisers cannot say good fortune confirms correctness when no check was made. Affected people need not consider every adverse event proof of malice. Both require description. Was the loss foreseeable? Was a way to reduce it available? Was an accepted promise broken? Which actions could actually have changed events? The questions guarantee no agreement. They limit disagreement to a factual subject rather than a competition between good intentions and the intensity of feeling. Moral clarity can include remaining uncertainty when information is insufficient.

Keep your share distinct from others' shares too. One participant might circulate an old poster after rescheduling was announced; another might promise to warn Mikhail personally and forget. Such circumstances change the discussion without automatically removing every question about Tamara. Perhaps notification was clearly delegated and accepted; perhaps it remained undefined. Delegating a task requires an intelligible subject and actual acceptance. Someone's silence does not always mean acceptance. An unclear distribution cannot later be presented as a precise agreement. Equally, real acceptance of someone else's task should not be erased to produce a single guilty person.

Chance does not always call for moral condemnation, but may call for changed practice. Even after a reasonable initial announcement, a paper poster proved longer-lived than the main notice. Organisers could add a last-confirmation date, a contact for checking or a message at the former venue. These need not acknowledge previous fault. They use new knowledge and an opportunity to reduce repetition. Change should not be confused with admitting that the previous choice was knowingly wrong. Otherwise people defend the past by refusing to learn. An honest assessment of earlier conditions can coexist with using discoveries for more attentive future decisions.

VIII. A boundary of responsibility must allow action

Responsibility for unintended losses is not ownership of the subsequent world. Rescheduling may affect dozens of decisions Tamara neither knows nor determines. Unlimited accountability for each would make any shared project impossible or produce an empty promise. Yet a reasonable boundary cannot consist only of “I did not want this.” Role, knowledge, available care, accepted arrangements and one's participation define it. Discovering a specific consequence adds possibilities for responding and reducing continuing loss. These subjects differ, but together give the boundary content.

In our scene Tamara may answer for inadequately updating known channels without answering for every later argument Mikhail has about the lost day. She can offer restoration of a particular loss without undertaking to organise all his future journeys. Mikhail can demand a response to a broken known arrangement without receiving unlimited authority over the organisation. The boundary should not be a unilateral declaration of the most convenient limit. Discuss it through actions and grounds; preserve the exact subject where disagreement remains. Any association rules are considered separately. The article does not settle authority or formal responsibility in their place.

A mature response can acknowledge a limit: I cannot restore your lost time now and do not know how many received the old poster. This is more honest than promising complete repair. But an available step should follow: contact distributors, correct the notice at the venue, clarify a specific expense with Mikhail or arrange further discussion. Not every limit can be removed; not every step suffices. A responsible boundary keeps a connection between what someone cannot do and what remains within their power. It makes uncertainty no convenient reason to do nothing further and requires no impossible knowledge as the price of participation.

Good intention remains significant when it withstands encountering consequences. It explains purpose without replacing examination of means. Foresight aids preparation without automatically making others' losses acceptable. Due care can reduce available risks without guaranteeing faultless results. Repair can restore part of a loss without rewriting history. These distinctions allow responsibility for one's share without appropriating the entire causal chain or abandoning it altogether. The book exchange's benefit survives an acknowledged error; precise responsibility makes shared work capable of considering people whose losses would otherwise remain outside a successful report.

What this means in practice

Choose an ordinary adult decision with a good purpose and possible inconvenience to others. Use this exercise neither as a legal assessment nor as a replacement for specific professional rules. Record purpose, chosen action and expected consequences separately. Identify which consequence is needed for the method and which is unnecessary for its success. Calling something incidental establishes neither insignificance nor acceptability. If the method cannot yet be described concretely, begin with actions, times, recipients and resources; assessment follows description.

Make three short lists: what is established before deciding, what could reasonably be checked and what emerged only afterwards. Put no later information into the past, and equate no ignorance with complete absence of fault. Give each available question an action: check an announcement channel, ask recipients, clarify dependence on the date or confirm an accepted task. Match the list to your role and time. Require no knowledge of every personal plan, but omit no readily available information about your own arrangements. Mark unestablished circumstances as such instead of filling them with convenient assumptions.

Consider at least the available alternatives and their genuine costs. What follows from delay, reduced scale, another notification method or abandoning part of the action? Does a person or group repeatedly carry costs for general success? How far can they join prior discussion? Promise no identical outcome, and treat no lack of complete safety as grounds for checking nothing. Define revision conditions: which new information can change the decision, and until when? If information can only accumulate without ever affecting the plan, examination remains formal.

If losses are discovered, first establish the consequence and your own participation. Acknowledgement requires no advance acceptance of every claimed obligation; it requires not denying the case through an appeal to good purpose. Discuss responsibility, continuing risk and available restoration separately. Check whether others can still encounter the same situation. A notice at the former venue may be more useful than prolonged debate about intentions. Clarify which part is reversible and which is not; promise no complete erasure. Applicable external rules and specific obligations are considered separately.

When distributing further work, name the task, recipient, deadline and confirmation method. Delegation does not end with assuming someone understood. If someone accepted a specific obligation and failed to fulfil it, that share remains theirs without necessarily cancelling yours. Keep a general causal chain distinct from concrete available responsibility. Replace no restoration with demanded immediate forgiveness. Those affected may recognise effort while remaining dissatisfied. Disagreement can stay precisely described: agreed facts, disputed obligations and proposed steps. This is preferable to unfounded declarations that every consequence has been removed.

After available steps are fulfilled, check actual conditions: is information corrected, are accepted tasks completed, is continuing loss reduced and what has been learned for the next decision? Changed practice does not necessarily prove earlier fault; it can use new knowledge. The exercise has no established efficacy and guarantees no absence of new losses. Its subject is a more accurate description of your share, enabling action without complete justification or boundless self-accusation. Good intention, due care and response to consequences are assessed separately; none should silently replace the others.

Key quotation

without due care

Thomas Aquinas, Summa Theologiae, II–II, question 64, article 8, fragment of the main answer. Fathers of the English Dominican Province translation. The fragment belongs to a historical discussion of accidentally causing death; no modern legal rules are derived from it here.

Connections between traditions

Question	Source	Distinction retained
What does someone know about their action's circumstances?	Aristotle, Nicomachean Ethics, III.1	Voluntariness, mixed actions and ignorance of particulars, with its own distinctions concerning repentance.
Does good intention remove proportion's requirement?	Thomas Aquinas, Summa Theologiae, II–II, 64.7	Intended and other effects, proportion within a specific historical theological subject.
Does chance remove the question of care?	Thomas Aquinas, II–II, 64.8	Due care and accidental harm; no modern legal test.
How should a discovered loss be answered?	Editorial comparison of the two sources	The article develops the distinction between earlier choice and present restoration; four rows represent no four traditions.

Sources

Aristotle. Nicomachean Ethics, Book III. W. D. Ross's translation, Internet Classics Archive, MIT; the translator appears at displayed line 21. Continuous III.1 was read at lines 26–51. The epigraph is eight English words from the discussion's concluding fragment at line 36; the Russian translation is our own. No reading of the entire Book III or work is claimed. Historical distinctions of voluntariness supply no modern risk-assessment procedure or legal determination of fault. Worldwide translation rights are unestablished; raw import is not authorised.
Thomas Aquinas. Summa Theologiae, II–II, question 64. Second and Revised Edition, 1920, translated by Fathers of the English Dominican Province; publication details were read at lines 141–144, with the online edition identifying copyright 2017 Kevin Knight. Continuous articles 7–8 were read, including objections, main answers and replies, at lines 107–140. The key fragment is three English words from the main answer to 64.8 at line 136; the Russian translation is our own. The source concerns self-defence, accidentally causing death and theological responsibility; no modern legal recommendation is derived. The online edition is copyrighted; worldwide translation status is unestablished, and raw import and indexing are not authorised.

About the authors

Andrey Zarinsh is the author and founder of United Wisdom. He explores inner transformation, the unity of 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.