{
  "id": "doctrine-of-double-effect",
  "title": "Doctrine of Double Effect",
  "lang": "en",
  "source_lang": "en",
  "summary": "The doctrine of double effect distinguishes intended harm from foreseen side effects when assessing whether an action with both good and harmful consequences is morally permissible.",
  "category": "philosophy",
  "subcategory": "ethics",
  "type": "concept",
  "aliases": [
    "Principle of double effect",
    "Rule of double effect",
    "Double-effect reasoning"
  ],
  "generated_at": "2026-11-04",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/doctrine-of-double-effect/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/doctrine-of-double-effect/",
      "json": "https://aiwiki.page/zh/doctrine-of-double-effect.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "The doctrine of double effect is a principle in ethics according to which an action producing both a good effect and a harmful effect may sometimes be morally permissible when the harm is foreseen but not intended, either as an end or as a means. Permission also depends on further conditions concerning the action’s moral character and the proportionality of the harm. The doctrine therefore does not hold that good intentions excuse any harmful conduct. It addresses circumstances in which pursuing a legitimate good involves harm that would ordinarily count against acting."
    },
    {
      "heading": "Historical development",
      "text": "The doctrine’s standard historical starting point is Thomas Aquinas’s discussion of self-defence in the Summa theologiae, II–II, question 64, article 7. Aquinas argues that a single act can have two effects: preserving the defender’s life and causing the aggressor’s death. He distinguishes the intended preservation of life from a death that occurs beyond the defender’s intention, while also requiring that defensive force remain proportionate.\n\nThis discussion is an antecedent of the doctrine rather than a statement of the complete four-condition formula familiar from later accounts. Subsequent writers in scholasticism and Catholic moral theology systematized the conditions for actions with mixed effects.\n\nThe doctrine became important within the natural-law tradition and the moral teaching of the Catholic Church. Its modern philosophical discussion extends beyond religious ethics, particularly into applied ethics, theories of action, and debates about harming people to protect others. Acceptance of the intention–foresight distinction does not require acceptance of its theological origins."
    },
    {
      "heading": "Standard conditions",
      "text": "Although formulations differ, a common version specifies four requirements:\n\n- The action must be morally good or morally neutral in its relevant character. An independently impermissible act is not made permissible merely by producing beneficial consequences.\n- The harmful effect must not be intended. It may be foreseen and accepted as a side effect, but it must not be chosen as an objective.\n- The harmful effect must not be the means of achieving the good effect. The benefit cannot be obtained through the very harm whose permissibility is in question.\n- There must be a proportionately serious reason for permitting the harmful effect. The good pursued must justify accepting the harm in the circumstances.\n\nMany formulations also emphasize avoiding unnecessary harm and considering less harmful alternatives. These requirements are sometimes treated as part of proportionality rather than as additional conditions.\n\nThe conditions work together. A good purpose alone is insufficient, and merely classifying harm as unintended does not establish permission. A foreseeable side effect may still be excessive, avoidable, reckless, or otherwise unjustified.\n\nProportionality is not simple aggregation\n\nProportionality requires attention to the seriousness and likelihood of harm, the importance of the good pursued, and the available alternatives. It is not necessarily a numerical calculation in which any sufficiently large benefit cancels any harm.\n\nIn traditional formulations, the prohibition on intending certain harms operates as a constraint on what may be done. This distinguishes the doctrine from forms of consequentialism that assess conduct solely by its consequences, and from versions of utilitarianism that require maximizing aggregate welfare. The doctrine is commonly associated with deontological ethics, although it also assigns consequences an indispensable role in moral assessment."
    },
    {
      "heading": "Intention, foresight, and means",
      "text": "The central distinction concerns the place of a harmful outcome in an agent’s practical plan.\n\nAn outcome is intended as an end when producing it is itself an objective. It is intended as a means when the agent chooses it as a way of accomplishing another objective. An outcome is merely foreseen when the agent anticipates it without adopting it in either role.\n\nWanting to save lives, for example, does not establish that every harmful step in a rescue plan is unintended. If the plan uses someone’s injury or death to secure the rescue, that harm may be intended as a means even if the agent deeply regrets it. Intention is therefore not identical to desire, pleasure, or approval.\n\nConversely, certainty that a harmful effect will occur does not, on the doctrine’s account, automatically establish that it is intended. Foreseeability concerns what the agent expects; intention concerns what the agent is trying to accomplish. Whether this distinction remains morally decisive when harm is certain is disputed.\n\nCausal structure and description\n\nThe doctrine requires an account of causation: does the beneficial result occur through the harm, or do benefit and harm arise through distinct causal pathways? Chronological order alone cannot settle this. An injury occurring before a benefit need not be its means, while an outcome occurring later may still form part of the agent’s plan.\n\nA related difficulty concerns how actions are described. The same conduct may be described narrowly as “moving an object” or more fully as “using a person’s body to stop a threat.” The first condition cannot be applied reliably if the description leaves out morally relevant features.\n\nDefenders therefore generally deny that agents can satisfy the doctrine simply by declaring a benevolent intention. What they choose, how their plan works, and what outcomes their success depends upon also matter."
    },
    {
      "heading": "Illustrative cases and applications",
      "text": "Self-defence\n\nIn Aquinas’s example, proportionate defensive action can aim at stopping an attack and preserving life, even though the aggressor’s death is foreseen. Excessive force is not justified merely because the defender’s ultimate purpose is survival.\n\nThis account distinguishes stopping a threat from choosing death as an objective. Applying that distinction to particular cases can be difficult, especially when the defensive method is highly likely to be lethal. The philosophical principle does not by itself determine the requirements of any jurisdiction’s law.\n\nThe trolley problem\n\nThe trolley problem is a prominent thought experiment used to examine the doctrine. In a familiar version, diverting a runaway trolley saves five people but causes it to strike one person on another track. One double-effect interpretation treats the single death as a side effect of redirecting the threat rather than as the means of saving the five.\n\nBy contrast, pushing someone into the trolley’s path so that the person’s body stops it appears to use harm as a means. The doctrine can thus explain a moral distinction between cases with the same number of deaths.\n\nThese interpretations are contested. Variants in which the trolley’s route loops back, or in which someone’s presence is necessary to stop it, test how causal structure and intention affect the judgment. The doctrine supplies a framework for analysing such cases, not an uncontested solution to them.\n\nMedical ethics\n\nDouble-effect reasoning appears in discussions of medical ethics, particularly symptom relief near the end of life. Its traditional application distinguishes treatment intended to relieve suffering, with a foreseeable risk of serious harm, from an intervention intended to cause death.\n\nThe ethical example must be separated from empirical claims about treatment. The doctrine does not establish that appropriately administered pain relief necessarily shortens life. Whether an intervention carries a particular risk is a clinical question, not something settled by moral reasoning alone.\n\nNor does the doctrine replace requirements concerning professional standards, informed consent, or patient autonomy. An intervention’s permissibility cannot be inferred solely from a clinician’s stated intention.\n\nArmed conflict\n\nIn debates about just war theory, double-effect reasoning is used to distinguish attacks directed at military objectives that foreseeably harm civilians from attacks that target civilians as an objective or use their suffering as a means.\n\nThe distinction does not make all incidental civilian harm permissible. Proportionality and precautions remain central, and the doctrine’s moral categories should not be treated as identical to the technical rules of international law."
    },
    {
      "heading": "Principal objections",
      "text": "Uncertainty about intention\n\nCritics question whether intention can be identified clearly enough to bear the doctrine’s moral weight. Human plans can involve mixed purposes, conditional objectives, and outcomes that agents neither positively desire nor regard as incidental.\n\nThere is also a risk of strategic redescription: an agent may characterize harm as a side effect while relying on it to accomplish the plan. Defenders respond by analysing the structure of the action rather than accepting self-reports at face value. Disagreement remains about which account of intention is appropriate.\n\nThe moral significance of the distinction\n\nEven where intention and foresight can be distinguished, critics ask why the difference should change permissibility when victims suffer the same harm. From some consequentialist perspectives, foreseeable injury demands justification regardless of its role in the agent’s plan.\n\nDefenders argue that deliberately making someone’s harm part of a project can constitute a distinctive wrong. The issue is not simply whether the agent has unpleasant attitudes, but whether the action treats another person’s injury as something to be achieved or used.\n\nA further question is whether intention changes an act’s permissibility, the agent’s moral responsibility, or both. A malicious purpose might make an agent more blameworthy without necessarily changing what action the circumstances require.\n\nIndeterminacy of proportionality\n\nThe proportionality condition may leave difficult questions unresolved. Different judgments about risk, necessity, alternatives, and the importance of competing goods can yield different conclusions even when all parties agree that the harm is unintended.\n\nConsequently, the doctrine is not a complete decision procedure. It depends on broader judgments about rights, duties, morally permissible action, and the significance of particular harms."
    },
    {
      "heading": "Distinction from related principles",
      "text": "The doctrine differs from the distinction between doing and allowing harm. An agent can actively cause an unintended harmful side effect, while an omission can be deliberately chosen to bring about harm. Causing harm is therefore not equivalent to intending it, and allowing harm is not equivalent to merely foreseeing it.\n\nIt also differs from the claim that “the end justifies the means.” Its means condition expressly denies that a beneficial objective automatically licenses harmful methods. Its distinctive claim is narrower: under additional constraints, accepting harm as a side effect can be permissible when choosing that harm as an end or means would not be."
    }
  ],
  "links": [
    "ethics",
    "thomas-aquinas",
    "scholasticism",
    "natural-law",
    "catholic-church",
    "applied-ethics",
    "consequentialism",
    "utilitarianism",
    "deontology",
    "causation",
    "law",
    "trolley-problem",
    "thought-experiment",
    "medicine",
    "informed-consent",
    "patient-autonomy",
    "international-law"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}