A compass for disciplined moral judgment
The real dispute is not kindness versus survival
A provocative title can make a complicated moral problem appear simple. Suicidal Empathy: Dying to Be Kind, by Gad Saad, is publicly presented as an argument that excessive compassion has inverted political priorities. The publisher’s description says that “empathy in politics leads to civilizational collapse” and attacks what it calls “maladaptively irrational altruism.”1
The strongest version of the concern deserves a serious hearing. Human beings can become absorbed by one vivid case and neglect many less visible people. We can confuse feeling intensely with judging accurately. Concern for an identifiable person can collide with fair procedures. Sympathy can be captured by propaganda, familiarity, beauty, proximity, or group loyalty. A public institution cannot distribute scarce goods solely according to who evokes the strongest emotion.
Yet none of this establishes that empathy is a civilizational disease, still less that moral seriousness requires indifference toward selected people. The phrase suicidal empathy combines several questions that must be separated:
- What happened, and who is responsible?
- What is another person thinking or experiencing?
- Do we care about that person’s welfare?
- What rules are fair to everyone?
- What action will reduce harm without violating rights?
Affective resonance may answer part of the second or third question. It cannot answer all five. But switching it off does not answer them either. A moral system needs more architecture than a single emotional dial.
The choice is not maximum empathy or zero empathy.
It is ungoverned emotion or disciplined moral judgment.
What the critique of empathy gets right
The Stanford Encyclopedia of Philosophy does not offer one official “Stanford view.” It is a continually maintained scholarly reference work whose entries survey disputes across philosophy and related sciences.2 Its 2025 revision of the empathy entry gives considerable ground to critics: empathy is not automatically moral, may be partial, and can conflict with justice.3
Empathic concern can produce unfairness
In two experiments published in 1995, C. Daniel Batson and colleagues found that empathy-induced concern could lead participants to give a particular person preferential access to a scarce benefit, violating an allocation rule they otherwise regarded as fair.4 The study does not show that empathy normally causes injustice, but it demonstrates the logical point: a motive to help one person is distinct from a motive to apply a general standard.
Batson later summarized decades of work by distinguishing altruistic motivation from moral motivation. Both have strengths and weaknesses. His conclusion was not to abolish care but that “wise orchestration is needed to harmonize their strengths.”5
Emotional resonance can bias judgment
Jean Decety and Jason Cowell’s review of developmental, behavioral, and neuroscience research describes the relation between empathy and morality as complex: empathy sometimes guides moral judgment and sometimes interferes with it.6 Emotional intensity is sensitive to salience and similarity, while justice typically asks us to apply reasons across cases - including cases that do not move us.
Empathic distress can exhaust the helper
Sharing another person’s suffering can become aversive and self-focused. Tania Singer and Olga Klimecki distinguish this from compassion: “Compassion does not mean sharing the suffering of the other.” It involves concern and a motivation to help without reproducing the other person’s pain inside oneself.7 Experimental work on compassion training suggests that this distinction may matter for resilience: compassion practice can increase positive affect and may help people respond to suffering without collapsing into distress.8
That warning is important. It is also narrower than the political claim that empathy itself is an existential threat. The next step is to ask what, exactly, the word is being used to mean.
“Empathy” is not one thing
The SEP entry begins with a caution: empathy refers to a wide range of capacities, and different disciplines use the term differently.3 Treating them as a single faculty produces confusion. Four distinctions are especially useful.
- Feeling with: Affective empathy
- Experiencing an emotion that resembles another person’s emotion. It can create connection, but also distress, partiality, and emotional contagion.
- Understanding: Cognitive empathy
- Representing another person’s perspective, beliefs, constraints, or likely motives. It is a form of social understanding, not necessarily approval or affection.
- Caring for: Empathic concern / compassion
- Other-oriented concern and motivation to improve someone’s welfare. It can remain calm and does not require mirroring suffering.
- Overload: Personal distress
- Self-focused discomfort when witnessing another’s suffering. It may motivate escape rather than help and should not be confused with compassion.
A fifth category sits outside this family: moral judgment. It asks what is permissible, fair, deserved, effective, or required. Moral judgment may use information gained through perspective-taking and motivation supplied by compassion, but it also needs principles and evidence.
| Statement | What it asks for | What it does not imply |
|---|---|---|
| “Tell me what happened from your point of view.” | Cognitive perspective-taking | Agreement, forgiveness, or immunity |
| “Your suffering matters.” | Concern or compassion | That no competing interest matters |
| “I feel your fear.” | Affective resonance | That the feeling is accurate or policy-relevant |
| “The rule applies equally.” | Justice and impartiality | Emotional coldness |
| “You are accountable for this act.” | Responsibility | Denial of dignity or context |
This taxonomy exposes a recurrent mistake: a request to understand someone is treated as a demand to excuse them. But understanding has an epistemic function. It can establish who acted, what alternatives existed, what consequences were foreseeable, and which intervention is likely to work. Those are conditions for responsible judgment, not alternatives to it.
Why “zero empathy” is not the same as reason
Rejecting empathy can feel like clearing emotion from the field so that reason can operate. Psychologically, that picture is incomplete. Anger, disgust, fear, resentment, pride, and loyalty do not disappear when concern for an out-group is switched off. They may simply become less visible to the person experiencing them.
People often avoid empathy because it feels effortful
Across eleven studies involving 1,204 participants, C. Daryl Cameron and colleagues found a robust preference to avoid empathy when people could choose between empathic and detached tasks. Participants associated empathy with cognitive effort, aversion, and low efficacy. When researchers increased participants’ sense that they were capable of empathizing effectively, the avoidance diminished.9
This does not mean every refusal to empathize is lazy. It does mean that detachment should not be automatically credited as objectivity. Sometimes it is an avoidance strategy - a way to reduce cognitive and emotional work.
Cognitive empathy and concern can support justice
A study of 265 adults by Decety and Keith Yoder separated emotional empathy, cognitive empathy, and empathic concern. Cognitive empathy and concern predicted greater sensitivity to injustice toward others; emotional empathy did not.10 That pattern fits the central distinction: emotional mirroring is not enough, but understanding and other-oriented concern can help people notice unfairness that does not directly affect them.
Moral perception requires attention
Iris Murdoch’s moral philosophy treats attention as the disciplined effort to see another person and a situation more truthfully, with less projection by the ego. Moral failure often begins before a decision, in the description we choose: whether someone appears as a person with a history, a nuisance, a threat, a symbol, or an interchangeable member of a category.11
Attention is not sentimental indulgence. It is a method of correcting fantasy. A judge, investigator, doctor, or policymaker who refuses information about the people affected is not necessarily tougher; they may simply be reasoning from a thinner factual model.
Zero empathy can mean zero emotional resonance. It should never mean zero curiosity, zero evidence, or zero recognition of personhood.
A better moral architecture: care under impartial discipline
The Western philosophical tradition contains many criticisms of sentimentality. It also contains equally strong warnings about cruelty, arbitrary coercion, and treating persons as categories. Read together, the classics do not recommend maximum feeling or maximum hardness. They recommend calibration.
Aristotle: practical wisdom, not a single rule
In the Nicomachean Ethics, Aristotle says virtue involves feeling and acting “at the right times … towards the right people … in the right way.”12 The point is not bland moderation. Some actions are simply wrong; others require judgment responsive to particulars. Practical wisdom determines what the situation calls for.
Adam Smith: sympathy corrected by the impartial spectator
Smith did not believe spontaneous sympathy was sufficient. He asks us to examine ourselves from a more impartial standpoint: “We suppose ourselves the spectators of our own behaviour.”13 The imagined spectator checks both selfish indifference and partial concern. Smith’s solution to biased sympathy is not callousness but enlarged, disciplined perspective.
Kant: dignity limits what may be done to a person
Kant’s formulation is familiar because it captures a hard constraint: “treat humanity … as an end … never as means only.”14 A person can be held accountable, restrained, or punished while remaining a person rather than a tool for deterrence, a symbol of a group, or raw material for political messaging. The modern philosophical idea of dignity likewise denotes a basic status that belongs equally to persons and grounds duties and rights.15
Mill: coercion needs a harm-based justification
John Stuart Mill’s harm principle states that coercive power is justified to prevent harm to others, not merely because conduct is disliked, unconventional, or thought harmful to the person doing it.16 Mill does not prohibit enforcement; he requires precision about the harm, the actor, and the reason for intervention.
Shklar: keep cruelty visible
Judith Shklar’s political ethic urges us to “put cruelty first” - to treat the deliberate infliction of fear and suffering on weaker people as the first political vice to watch for.17 This does not imply permissiveness toward wrongdoing. It means that a rhetoric of necessity, purity, or survival should never make the cruelty of a proposed remedy disappear from view.
Perspective-taking supplies information.
Compassion supplies motivation.
Justice supplies constraints.
Practical wisdom integrates them.
The empirical case for complementarity
A 2026 study in PNAS Nexus examined effective altruists, extraordinary altruists such as non-directed organ donors, and comparison participants. Across many measures, both empathy and reasoning were associated with more equitable and effective altruistic attitudes and behavior; in several contexts the combination predicted the strongest outcomes.18 The study is correlational and uses distinctive populations, so it cannot settle causation. But it directly challenges the assumption that empathy and reason are natural enemies.
Evidence at a glance
- 1995 · EXPERIMENTS: Empathy can conflict with fair allocation
- Supports: concern for a salient individual can override a rule participants otherwise regard as just. Does not show: that concern usually produces injustice or that indifference improves allocation.
- 2014 · REVIEW: Empathy and morality have no simple relationship
- Supports: different affective, cognitive, and motivational components can help or hinder moral judgment. Does not show: that every component should be treated as one faculty.
- 2016 · N = 265: Understanding and concern can track justice
- Supports: cognitive empathy and empathic concern predicted sensitivity to injustice toward others. Does not show: that emotional resonance by itself improves judgment.
- 2019 · 11 STUDIES: Empathy avoidance can reflect perceived effort
- Supports: people often choose detachment when empathy feels difficult or ineffective. Does not show: that every refusal to engage is motivated by laziness or avoidance.
- 2012 · 2 CONFLICTS: Being heard can matter as much as perspective-taking
- Supports: dominant and non-dominant groups may need different dialogue roles. Does not show: that brief listening interventions resolve structural conflict.
- 2026 · MULTI-SAMPLE: Reason and empathy can be complementary
- Supports: both capacities were associated with equitable and effective altruism, often most strongly together. Does not show: causation; the samples were unusual and replication remains important.
Listening is not surrender - and power changes what helps
“Try to see it from their perspective” is not always the most effective instruction. Sometimes the person with less power has spent years understanding the dominant group and needs, instead, to be heard.
Emile Bruneau and Rebecca Saxe tested structured dialogue in two intergroup conflicts: White Americans and Mexican immigrants in Arizona, and Israelis and Palestinians. Perspective-taking was more beneficial for members of the dominant groups, while members of the less powerful groups responded more positively to perspective-giving - describing their experience and receiving evidence that it had been understood.19
The studies were limited in size and context, and some effects weakened with time. They do not prove that dialogue solves entrenched conflict. They do show why “just empathize” is too crude. Moral inquiry should ask who is routinely listened to, who already performs the work of understanding the other side, and what information is missing.
A responsible process therefore separates hearing from verdict. Investigators interview suspects without promising acquittal. Doctors listen to patients without granting every requested treatment. Courts permit testimony while still testing evidence. Listening is part of due process because judgment without information is brittle.
Individual responsibility is the barrier against collective blame
One of the most dangerous shortcuts in moral reasoning is replacing an act with an identity. “These people committed these acts” is an evidential claim. “This kind of people does this” turns a category into an agent.
Groups are not automatically moral agents
The SEP’s entry on collective responsibility distinguishes a unified collective agent - such as an organized institution capable of forming intentions - from an assortment of people who share a label. Philosophers disagree about when collective responsibility is justified, but they emphasize organization, joint action, contribution, and individual agency.20 An ethnicity, class, nationality, or social status does not by itself form a single acting mind.
Ordinary moral responsibility also contains epistemic conditions: blame normally depends on what an agent did, understood, could foresee, and could control.21 Category membership supplies none of those facts.
Dehumanization can occur without literally denying humanity
The SEP surveys several theories of dehumanization. Some focus on regarding people as less than human; others focus on practices that deny agency, individuality, subjectivity, or equal standing.22 Treating people as fungible examples of a threatening class can therefore be dehumanizing even when no one explicitly says they are not human.
Research by Nour Kteily and colleagues found that blatant dehumanization predicted aggressive intergroup attitudes beyond ordinary prejudice or dislike.23 Related research shows that collective blame - holding an entire group responsible for acts by some members - can license hostility. Interventions that expose inconsistent standards of collective blame have reduced such attitudes in experiments and produced effects that, though weakened, persisted in longer-term follow-up.2425
Three books that illuminate the slide
Martin Buber’s I and Thou distinguishes meeting another as a subject from treating them as an “It” - an object, function, or case.26 Hannah Arendt’s The Origins of Totalitarianism examines how propaganda, statelessness, and mass categorization can strip people of effective personhood and rights.27 Jonathan Glover’s Humanity traces the moral resources that restrain cruelty - and how tribalism, ideological distance, and diminished personal responsibility weaken those restraints.28
They differ profoundly, but converge on a warning: when identity becomes sufficient evidence, the distance between judgment and punishment collapses.
A practical decision checklist
A better decision procedure should preserve the valid concern about sentimental bias while blocking the drift into indifference, dehumanization, or collective punishment. The following five-stage checklist is designed for public-order conflicts, institutional decisions, and morally charged disputes.
- Clarify the conduct
Describe observable acts, evidence, frequency, scale, and harm. Do not substitute identity for behavior.
- Listen for facts
Hear affected residents, alleged offenders, victims, front-line workers, and relevant experts. Listening is inquiry, not acquittal.
- Evaluate impartially
Apply harm, rights, dignity, equality, due process, and proportionality to everyone involved.
- Act specifically
Target the conduct and responsible agents. Combine prevention, support, and enforcement where evidence warrants.
- Review outcomes
Measure recurrence, displacement, safety, costs, and unintended harms. Revise policies that fail.
Why this is harder than a slogan
The checklist denies both sides an easy moral performance. It denies the sentimentalist the right to ignore harms because the person causing them is vulnerable. It denies the punitive moralist the right to ignore facts, rights, or individual responsibility because the target is unpopular.
It also treats policy as an empirical practice. A measure is not rational merely because it is severe, just as it is not humane merely because it sounds compassionate. An intervention must survive evidence about what it actually does.
A worked example: suspected AI-assisted cheating
Suppose a university instructor receives an essay with fabricated citations, abrupt shifts in style, and a high score from an automated AI-writing detector. The course rules prohibit submitting generated prose as one’s own but allow tools for outlining and language correction. The university has legitimate interests in assessment integrity, fair treatment of students who did the work, and credible qualifications. The student has legitimate interests in clear rules, an evidence-based inquiry, due process, privacy, and a response proportionate to what can actually be established.
Failure mode A: permissive sentimentalism
- Treats stress, unequal preparation, or unclear technology as immunity from academic rules.
- Minimizes the harm to classmates who completed the work and to the value of the qualification.
- Accepts a sympathetic explanation without testing it against available evidence.
- Offers understanding while leaving students uncertain about what the institution will enforce.
Failure mode B: punitive essentialism
- Treats a detector score or unusual prose as proof rather than a reason to investigate.
- Treats every flagged submission as evidence of the same conduct and intent.
- Skips the student’s account, drafts, notes, and opportunity to explain the work.
- Imposes the maximum penalty before establishing the scope of the breach or whether the rule was clear.
Applying the checklist
- Clarify: state the alleged breach, the rule in force when the work was assigned, and the evidence that raised concern; distinguish prohibited generated prose from permitted outlining, editing, or language assistance.
- Listen: ask the student to explain the argument, sources, and writing process; examine relevant drafts, notes, or version history alongside the instructor’s observations without demanding unrelated private data.
- Evaluate: assess the evidence independently of the detector score, the clarity of the instructions, the consistency of standards across students, the apparent intent and extent of any misuse, and the student’s procedural rights.
- Act: if the evidence supports a breach, use an individual and proportionate response such as an oral defence, supervised rewrite, grade consequence, or formal discipline appropriate to its severity; if it does not, close the allegation without converting suspicion into a permanent label.
- Review: examine appeals, error patterns, differences across courses, and whether the assessment still measures the intended learning; revise the policy, teaching, or assessment design where recurring ambiguity appears.
The example shows why understanding context and maintaining standards are not opposing tasks. Stress, unclear rules, or permitted uses of AI may affect intent and the appropriate response, but they do not settle whether the submitted work is the student’s own. Conversely, suspicion and detector output may justify an inquiry, but not a finding of guilt. Compassion belongs in the quality of the investigation and the proportionality of the response; accountability belongs in the evidence, the rule, and the student’s own conduct.
This is not a compromise between being “soft” and being “hard.” It is a more demanding standard: be accurate, humane, lawful, and effective at the same time.
Ten books that deepen the argument
The following sequence moves from moral psychology to political restraint, personhood, historical dehumanization, and the modern critique of emotional empathy.
Adam Smith - The Theory of Moral Sentiments
Begin here for sympathy corrected by the impartial spectator. Smith refuses both selfish indifference and ungoverned partial feeling.
Foundation · Moderate · moral psychology
Aristotle - Nicomachean Ethics
Shows why good judgment cannot be reduced to maximizing one emotion. Virtue responds to particulars through cultivated practical wisdom.
Practical wisdom · Moderate · ancient ethics
Iris Murdoch - The Sovereignty of Good
A short, demanding account of attention, unselfing, and the moral work of seeing another person without fantasy.
Moral perception · Dense · three essays
Judith Shklar - Ordinary Vices
Examines cruelty, hypocrisy, betrayal, snobbery, and misanthropy. Essential for testing rhetoric that presents harshness as virtue.
Political restraint · Accessible · political ethics
Immanuel Kant - Groundwork of the Metaphysics of Morals
Provides the clearest classical limit on instrumentalizing people: each person must be treated as an end.
Dignity · Demanding · short classic
John Stuart Mill - On Liberty
Separates concrete harm from mere offense and supplies a disciplined test for when social or legal coercion is justified.
Coercion · Accessible · liberal theory
Martin Buber - I and Thou
Explores the difference between encountering a person as a subject and reducing them to an object, role, or category.
Relation · Poetic · short and abstract
Hannah Arendt - The Origins of Totalitarianism
Read especially the sections on statelessness, propaganda, and the loss of effective personhood when people become population problems.
Rights and exclusion · Long · historical-political
Jonathan Glover - Humanity
Studies how cruelty becomes possible and what moral resources - respect, sympathy, responsibility - can restrain it.
Moral history · Long but readable
Paul Bloom - Against Empathy
The strongest accessible case against emotional empathy, paired with a defense of rational compassion. Useful for distinguishing a careful critique from a culture-war slogan.
Modern counterpoint · Accessible · contemporary psychology
A compact three-book route is Smith for impartial sympathy, Murdoch for truthful attention, and Shklar for keeping cruelty visible. Add Bloom to see how a rigorous critic of empathy can defend compassion rather than indifference.29
Empathy is a sensor, not the steering wheel
The critique of empathy begins from a genuine problem. Emotional identification is selective. It can be manipulated, exhausted, and directed toward one person at the expense of fair rules or unseen others. Institutions need more than feeling.
But the remedy is not to make understanding morally suspect or to reserve concern for those already close to us. “Zero empathy” is not a neutral platform from which reason speaks. It can leave anger, disgust, fear, and group loyalty entirely intact while removing one source of corrective information.
The better tradition - from Aristotle and Smith to Murdoch, Shklar, Mill, Kant, and contemporary psychology - treats moral judgment as a coordinated achievement. We need perspective-taking to understand the situation, compassion to recognize welfare as morally salient, impartiality to correct favoritism, dignity and rights to constrain action, and practical wisdom to choose a proportionate response. Then we need evidence about whether it worked.
Compassion without justice can become favoritism. Justice without attention can become blindness. Enforcement without dignity can become cruelty.
The mature position is neither sentimental permissiveness nor proud callousness. It is disciplined humanity: see clearly, judge impartially, hold individuals accountable, protect everyone affected, and remain willing to revise the story when the facts change.
Sources
- Google Books / HarperCollins publisher description, Suicidal Empathy: Dying to Be Kind, 2026. Promotional description quoted sparingly.
- Stanford Encyclopedia of Philosophy and editorial guidelines. SEP is a scholarly reference project, not a single institutional position.
- Karsten Stueber, “Empathy,” Stanford Encyclopedia of Philosophy, substantive revision 6 October 2025.
- C. Daniel Batson, Tricia Klein, Lori Highberger, and Laura Shaw, “Immorality from Empathy-Induced Altruism: When Compassion and Justice Conflict”, Journal of Personality and Social Psychology 68, no. 6 (1995): 1042 - 1054.
- C. Daniel Batson, “Empathy-Induced Altruism and Morality: No Necessary Connection”, in Empathy and Morality (Oxford University Press, 2014).
- Jean Decety and Jason Cowell, “The Complex Relation Between Morality and Empathy”, Trends in Cognitive Sciences 18, no. 7 (2014): 337 - 339.
- Tania Singer and Olga M. Klimecki, “Empathy and Compassion”, Current Biology 24, no. 18 (2014): R875 - R878.
- Olga M. Klimecki et al., “Differential Pattern of Functional Brain Plasticity after Compassion and Empathy Training”, Social Cognitive and Affective Neuroscience 9, no. 6 (2014): 873 - 879.
- C. Daryl Cameron et al., “Empathy Is Hard Work: People Choose to Avoid Empathy Because of Its Cognitive Costs”, Journal of Experimental Psychology: General 148, no. 6 (2019): 962 - 976.
- Jean Decety and Keith J. Yoder, “Empathy and Motivation for Justice: Cognitive Empathy and Concern, but Not Emotional Empathy, Predict Sensitivity to Injustice for Others”, Social Neuroscience 11, no. 1 (2016): 1 - 14.
- Niklas Forsberg, “Iris Murdoch,” Stanford Encyclopedia of Philosophy; see also Iris Murdoch, The Sovereignty of Good.
- Aristotle, Nicomachean Ethics, Book II, trans. W. D. Ross.
- Adam Smith, The Theory of Moral Sentiments; see also SEP, “Adam Smith’s Moral and Political Philosophy”.
- Immanuel Kant, Fundamental Principles of the Metaphysic of Morals.
- Remy Debes, “Dignity,” Stanford Encyclopedia of Philosophy, 2023.
- John Stuart Mill, On Liberty.
- Judith N. Shklar, “Putting Cruelty First”; developed in Ordinary Vices (Harvard University Press, 1984).
- Kyle Fiore Law et al., “Reasoning and Empathy Are Not Competing but Complementary Features of Altruism”, PNAS Nexus 5, no. 2 (2026): pgag015.
- Emile G. Bruneau and Rebecca Saxe, “The Power of Being Heard: The Benefits of ‘Perspective-Giving’ in the Context of Intergroup Conflict”, Journal of Experimental Social Psychology 48, no. 4 (2012): 855 - 866. Author-accessible PDF.
- Marion Smiley, “Collective Responsibility,” Stanford Encyclopedia of Philosophy.
- Fernando Rudy-Hiller, “The Epistemic Condition for Moral Responsibility,” Stanford Encyclopedia of Philosophy; Michael McKenna and D. Justin Coates, “Moral Responsibility”.
- Bertrand Phillips, “Dehumanization,” Stanford Encyclopedia of Philosophy, 2025.
- Nour Kteily, Emile Bruneau, Adam Waytz, and Sarah Cotterill, “The Ascent of Man: Theoretical and Empirical Evidence for Blatant Dehumanization”, Journal of Personality and Social Psychology 109, no. 5 (2015): 901 - 931.
- Emile Bruneau, Nour Kteily, and Emily Falk, “Interventions Highlighting Hypocrisy Reduce Collective Blame of Muslims for Individual Acts of Violence and Assuage Anti-Muslim Hostility”, Personality and Social Psychology Bulletin 44, no. 3 (2018): 430 - 448.
- Emile G. Bruneau, Nour S. Kteily, and Ana Urbiola, “A Collective Blame Hypocrisy Intervention Enduringly Reduces Hostility Towards Muslims”, Nature Human Behaviour 4 (2020): 45 - 54.
- Martin Buber, I and Thou, trans. Walter Kaufmann.
- Hannah Arendt, The Origins of Totalitarianism.
- Jonathan Glover, Humanity: A Moral History of the Twentieth Century.
- Paul Bloom, Against Empathy: The Case for Rational Compassion.
Research findings are summarized at the level supported by the cited publications. Individual studies vary in design, sample, and generalizability; no single experiment settles a broad philosophical dispute.