Grossman, a professor of military science and psychology at West Point and later a career Army officer and researcher, spent years investigating this finding — questioning it, testing it, trying to understand it. Marshall’s methodology had been disputed. Some of his specific numbers were later challenged. But the underlying observation — that human beings have a profound, apparently innate resistance to killing other human beings — survived the challenge to the methodology. It appeared across multiple studies, in multiple contexts, in multiple historical periods. The resistance was real. Deep. And something the military had to systematically work around.
The story of how the military worked around it — and what the psychological cost of that workaround has been — is the central subject of Dave Grossman’s On Killing: The Psychological Cost of Learning to Kill in War and Society. This is a deep exploration of one of the most important books ever written about the psychology of violence, combat, and the human capacity for harm.
What follows covers Grossman’s analysis of the resistance to killing, the conditioning techniques the military developed to overcome it, the psychological aftermath of killing in combat, the relationship between combat PTSD and moral injury, and the implications of Grossman’s framework for how violence in broader society gets understood. Not comfortable material. Essential material.
The Resistance to Killing — An Innate Human Reality
Grossman’s central thesis is that human beings are uniquely and specifically resistant to killing other members of their own species — that this resistance is hardwired rather than merely culturally produced, that it’s consistent across cultures and historical periods, and that the evidence for it is visible in the historical record of combat in ways that have been systematically underacknowledged.
The evidence he marshals is extensive. Beyond Marshall’s WWII data, Grossman examines the actual casualty rates of historical battles and compares them with what they should have been if all soldiers with weapons had used them against visible enemies. He finds, repeatedly, a massive discrepancy: casualty rates that make sense only if a large proportion of soldiers were firing in directions that reduced the probability of hitting anyone, or not firing at all.
He examines the accounts of soldiers across multiple wars who describe having enemy fighters in their sights and being unable to pull the trigger. Not from cowardice — many of these same men performed extraordinary acts of physical bravery under fire. From something different: a specific inability to execute the act of killing a person who was visible to them as a person. The anonymized target is one thing. The young man looking back across a field is another, and something in the nervous system treats them differently.
Grossman uses the concept of “empathy” broadly to describe this resistance — the intrinsic capacity to imaginatively inhabit the experience of another person, which makes killing that person feel like a violation of something fundamental in the self. Not sentimentality. A description of a neurological reality: mirror neurons, social cognition systems, and the whole apparatus of human empathy that evolved in the service of social cohesion now creates a barrier against lethal violence against in-group members — a barrier so strong that even the extreme stress and danger of combat does not reliably override it.
“There is a powerful, innate human resistance to killing your fellow man. A resistance so strong that, in many circumstances, soldiers on the battlefield would die rather than overcome it.” — Dave Grossman, On Killing
Overcoming the Resistance — The Military Conditioning Program
After World War II, the finding that only fifteen to twenty percent of riflemen were consistently engaging the enemy drove a systematic redesign of military combat training. The goal was not to make soldiers braver — the evidence suggested bravery was not the limiting factor. The goal was to condition soldiers to overcome the specific, innate resistance to killing. The methods that were developed, which Grossman documents in detail, are sophisticated, effective, and represent some of the most consequential applied psychology of the twentieth century.
The first major change was desensitization — reducing the psychological impact of killing through repeated, normalized exposure to the act in training. The introduction of human-silhouette targets, replacing the bulls-eye targets that had been standard, was an early step. A bulls-eye target provides no social cue — pure geometric shape. A human silhouette triggers the social recognition systems that generate the killing resistance. Training soldiers to shoot at human-shaped targets repeatedly, under the operant conditioning framework of immediate reinforcement (the target falls when hit), begins to recode the association between human shape and the act of shooting.
The second change was conditioning — creating automatic, reflexive responses to combat stimuli through intensive, realistic training. The behavioral psychology principle is straightforward: if the training environment is sufficiently similar to the combat environment, the responses conditioned in training will transfer to combat. The more realistic the training, the more completely the conditioned response overrides the resistance. Modern military training has pushed this further with scenario-based training, live-fire exercises, and simulation technology that approximates the genuine sensory environment of combat as closely as possible.
The results were dramatic. Studies of WWII versus Korean War versus Vietnam War firing rates showed consistent increases from the fifteen to twenty percent of WWII to figures approaching ninety percent in Vietnam. The conditioning program had essentially solved the military’s problem. It had overcome the innate resistance to killing.
But the success of the conditioning came with a consequence the military had not fully anticipated, and that Grossman argues has never been fully reckoned with.
The Cost — What Happens to Men Who Kill
The conditioning that makes soldiers capable of killing in combat does not resolve the psychological resistance that was overcome. It bypasses it. The resistance does not disappear. And after combat, when the training stimulus is no longer present and the soldier is no longer in a life-threatening environment that justifies killing, the bypassed resistance reasserts itself — often with devastating force.
Grossman documents the psychological aftermath of killing in combat in clinical detail, drawing on accounts from soldiers, clinical literature on combat PTSD, and his own extensive professional experience. What he describes is a specific form of psychological injury distinct from fear-based PTSD — distinct from the hypervigilance and intrusive memories that come from exposure to life threat. It’s the injury that comes from having done something that violates one’s deepest moral self-concept. Today this would be called moral injury, a term that wasn’t widely used when Grossman wrote the book but that has become the standard clinical descriptor for exactly the phenomenon he’s describing.
Moral injury, as articulated by Jonathan Shay, Grossman, and subsequent researchers, is the damage done to the moral self by one’s own actions or the actions of others that violate deeply held moral beliefs. Different from PTSD in its mechanism and in its treatment requirements, though the two often co-occur in combat veterans. PTSD is anchored in the fear response. Moral injury is anchored in the sense of having transgressed something fundamental.
For soldiers who killed in combat — particularly those who killed at close range, where the human reality of the person killed was most visible and most impactful — the return to civilian life often brought a confrontation with that transgression that was genuinely unbearable. Intrusive imagery not of things done to them but of things done by them. The face of the person they killed. The moment the resistance was overcome and the trigger was pulled. These images, Grossman argues, persist because the moral processing system cannot complete the transaction — cannot reach a resolution that accepts killing as consistent with the self that existed before combat.
Case Study — The Face That Wouldn’t Leave
One of the accounts Grossman includes, drawn from clinical and oral history sources, describes a soldier from the Vietnam era who killed an enemy combatant at close range during a night patrol ambush. He described the kill in clinical terms years later — the tactical moment, the decision, the outcome. And then he described what happened in the years after: the face. The face of the man he had killed, appearing in his sleep, appearing when he was driving, appearing during family meals, appearing at the most mundane and domestically innocent moments of civilian life, superimposed on normalcy like a photograph that wouldn’t fade.
He was not a broken man. He had a family, a career, a community. Not reckless or violent or dangerous. He carried this face with him as a private weight, something never told to anyone, something with no language in his civilian world. He brought it eventually to a veterans’ group, where he found, for the first time, other men who understood the specific content of what he was carrying. Not the fear of combat. Not the chaos of the battlefield. The specific weight of a face.
Grossman uses accounts like this one to make a specific point: the killing resistance that was overcome in combat does not stop having moral weight when the killing was justified, lawful, and necessary. The psychological system that generates the resistance does not operate on those categories. It registers the act at a level below justification. And the processing work — the moral and psychological work of integrating that act into a coherent sense of self that can carry both the act and the person who performed it — is work that many veterans have done alone and in silence because their culture had no framework for supporting it.
The Spectrum of Responses — Who Suffers and Why
Grossman does not claim that all soldiers who kill in combat develop severe psychological problems — the data doesn’t support that claim and he doesn’t make it. What the research shows is a spectrum of responses to killing, ranging from minimal lasting impact through significant adjustment difficulties through severe and chronic moral injury and PTSD, and that the position on that spectrum is determined by a set of identifiable factors.
The most important predictors of severe impact, Grossman identifies, are proximity of kill (close-range killing has dramatically more psychological impact than distant or anonymous killing), perceived legitimacy of the killing (kills that the soldier retrospectively doubts were justified — innocent civilians, prisoners, kills under morally ambiguous circumstances — produce more lasting damage than kills that remain clearly justified in retrospect), social support during and after (the presence of unit cohesion during combat and genuine social reintegration after it are among the most powerful buffers against lasting damage), and whether the soldier had an existing moral framework capacious enough to accommodate having killed.
That last factor deserves attention. Grossman, drawing on the military chaplaincy tradition and the historical role of religious frameworks in preparing warriors, argues that men who go to war with a coherent moral framework that includes a concept of justified killing — that distinguishes killing in defense of others from murder, that provides language and ritual for processing the act — are substantially better protected against lasting psychological damage than men who go to war without such a framework and who are then left to construct one retroactively from scratch. Not an argument for any particular religious tradition. An observation about the function of moral frameworks in providing the cognitive tools that moral integration requires.
The Vietnam Effect — Institutional Failure and Its Consequences
Grossman’s chapter on Vietnam is among the most important in the book, and it examines the specific ways in which the institutional context of the Vietnam War amplified the psychological costs of combat killing in ways that were not inevitable but that reflected specific policy failures.
The first factor was the individual rotation system. Unlike WWII, where soldiers served with the same unit for the duration of their deployment, Vietnam used a twelve-month individual rotation that sent soldiers to their units at different times, broke up unit cohesion, and ensured that the social support structure that is among the most important buffers against psychological damage was perpetually disrupted. A soldier who arrived in a unit as a replacement, who built relationships and then rotated home while his unit remained — or who watched his unit fragment as men were rotated, wounded, or killed — never developed the deep unit bonds that research consistently identifies as protective.
The second factor was the homecoming. Vietnam veterans returned individually, not in units, directly from combat to civilian airports. Many encountered active hostility from anti-war demonstrators. None received the decompression period, the unit reintegration rituals, or the community celebration that historically have served the function of transitioning warriors from combat to civilian identity. They went from firefights to suburban living rooms with nothing in between. The psychological transition that requires time, ritual, and community support received none of these.
The consequences are documented in the epidemiology: Vietnam veterans showed dramatically higher rates of PTSD, substance abuse, marital disruption, incarceration, and suicide than veterans of comparable combat exposure in other wars. These are not primarily consequences of the intensity of the combat. They’re consequences of the failure of the institutional container that should have supported the transition from warrior to citizen.
This matters beyond its historical specificity, because the lessons apply to every institutional context in which people are asked to do psychologically difficult things and then expected to manage the aftermath alone. The research Grossman documents supports what Sebastian Junger argues in Tribe and what the broader social support literature consistently shows: human beings are not designed to process extreme experiences alone. The absence of genuine community support does not mean the processing gets done independently. It means it doesn’t get done.
Violence in Media — The Conditioning Effect on Civilians
The final major section of Grossman’s book is the most controversial and the one that has generated the most critical response: his argument that the same conditioning techniques used to train soldiers to overcome the killing resistance are being deployed, inadvertently or deliberately, through violent video games and media, to a civilian population — including children — who have none of the screening, ethical training, or institutional context of military conditioning.
Grossman argues that the repetitive, realistic, rewarded simulation of killing in first-person shooter games activates the same conditioning mechanisms military trainers use intentionally: desensitization through exposure, operant conditioning through reward (the kill counter, the score, the advancing game level), and the normalization of the act through the framework of entertainment. He draws an explicit parallel between the human-silhouette targets military training introduced and the photo-realistic human targets of modern first-person games.
This section of the book is the most scientifically contested. The research on video game violence and real-world aggression is genuinely mixed — some published data shows effects, others don’t, and the magnitude of any causal relationship is debated. Grossman’s position is stronger on the desensitization side than on the direct aggression side: the evidence that exposure to repeated realistic simulations of violence reduces the affective response to violence is stronger than the evidence that it increases the probability of violent behavior. Different claims, and conflating them has contributed to the debate being less productive than it should be.
What Grossman gets right, regardless of the specific video game debate, is the principle: conditioning techniques that can be used to train people to do things they would not naturally do are not morally neutral entertainment. The same mechanism that overcame the killing resistance in Vietnam-era soldiers can be engaged by civilian entertainment. Whether the threshold effect is as dramatic as Grossman argues is an empirical question. That the effect exists at all is not seriously disputed.
The practical takeaway
On Killing is not a book about whether war is just or whether violence is ever warranted. It’s a book about what violence costs the people who practice it — even when it’s necessary, even when it’s heroic, even when it’s fully justified by every standard available. That cost is real, it’s physiologically instantiated, and it does not diminish because the killing was right.
The broader lesson — the one that extends well beyond the military context — is that the psychological systems that make us human and social are not bypassed without consequence, and that the most important thing a society can do for the people it asks to do hard and morally complex things is to prepare them honestly for what those things will cost, and to provide genuine support for the integration work afterward.
Vietnam veterans were failed spectacularly on both counts. The measure of whether we do better going forward is whether the knowledge gets taken seriously enough to act on. Grossman provided the knowledge. What gets done with it is the ongoing responsibility. This connects directly to Sebastian Junger’s Tribe, which takes these questions about belonging and veteran reintegration even further. The resilience framework and mindset tools provide supporting context for everything covered here. Stay hard, stay honest, stay in the work.
Key Lessons from On Killing
- The resistance to killing is real, hardwired, and consistent across cultures. Only 15-20% of WWII riflemen consistently fired at the enemy. This is not cowardice — it is a deep, apparently innate resistance to killing other human beings that appears across the historical record.
- The military solved the problem by bypassing the resistance, not eliminating it. Modern conditioning techniques raised firing rates to 90%+ in Vietnam. The resistance didn’t disappear — it was overridden. The cost of that override is what creates moral injury.
- Moral injury is different from PTSD and requires different treatment. It is the damage done to the moral self by transgressing deeply held values — not a fear-based disorder but an ethics-based one. Treatment requires genuine moral and community support, not just symptom management.
- The Vietnam homecoming failure was preventable and catastrophic. Individual rotation, hostile homecomings, and the absence of genuine community reintegration rituals produced dramatically worse psychological outcomes than comparable wars with better institutional support.
- Social support is the most powerful buffer against lasting damage. Unit cohesion during combat and genuine community reintegration after it are among the most effective protective factors — more effective than individual resilience traits alone.
Straight Talk on On Killing
On Killing is a foundational work in the psychology of violence, combat, and the human capacity for harm. It is not an antiwar book, though it documents the enormous human cost of war. It is not a pro-war book, though it treats warriors with genuine respect. It is an honest reckoning with what it costs human beings to do the thing that war requires — and with the ways that those costs can be minimized through better preparation, better institutional support, and more honest public engagement with what military service actually involves. Anyone who cares about either military personnel or the honest understanding of violence should read it.
Books Similar to On Killing
The following books extend and deepen Grossman’s framework. Meditations on Violence by Rory Miller provides the practitioner’s perspective on real-world violence from a corrections officer’s experience. Tribe by Sebastian Junger covers the belonging and reintegration dimension that Grossman identifies as critical. Achilles in Vietnam by Jonathan Shay develops the moral injury concept from a clinical perspective with deep engagement with Homer’s treatment of warrior psychology. Soft Wired by Michael Merzenich provides neuroscientific context for the plasticity and conditioning mechanisms Grossman describes. The Body Keeps the Score by Bessel van der Kolk covers the physiological dimensions of trauma processing that complement Grossman’s psychological framework.
Who Should Read On Killing
Anyone with a genuine interest in understanding what combat does to human beings needs this book. Veterans who are trying to make sense of their own experience will find a framework that treats their psychology with the seriousness it deserves. Mental health professionals working with combat veterans need the moral injury framework. Policymakers and citizens who make decisions about when and how to send soldiers to war need the honest accounting of what those decisions cost the people who execute them. Parents of military service members will understand their children’s experience more accurately and more compassionately with this knowledge. It is, in the end, a book about what it means to be human in extremity — which is relevant to everyone.
Practical Protocol — For Those Working With Veterans
Distinguish moral injury from PTSD. Moral injury and PTSD can co-occur and can look similar, but they are different in their mechanism and in their treatment requirements. The veteran who is experiencing intrusive images of what was done to them is primarily experiencing fear-based PTSD. The veteran who is experiencing intrusive images of what they did is primarily experiencing moral injury. These require different clinical approaches and different forms of community support.
Provide genuine moral framework support. Moral injury heals through the construction of a coherent moral account of the actions that caused it — an account that is honest about the reality of what happened and that can carry both the act and the person who performed it. This requires genuine moral engagement, not reassurance. “You were just following orders” is not a moral framework. “You did a terrible thing in service of something genuinely worth protecting” is the beginning of one.
Build genuine community before clinical intervention. The evidence from Grossman and the subsequent literature consistently shows that social support — genuine community with people who share the relevant experience — is among the most powerful therapeutic factors available. The veteran who is embedded in a genuine community of people who understand their experience is better positioned to do the clinical work than the veteran who is isolated. Community building is not a supplement to treatment; it is often a prerequisite for it.
Killing Summary: Your Questions Answered
Does Grossman’s killing resistance theory apply to women as well as men? Yes. The research on killing resistance does not show significant gender differences — the resistance appears to be a general human characteristic rather than a specifically male one. Female soldiers who kill in combat show similar moral injury patterns to male soldiers.
Is Grossman’s criticism of violent video games scientifically supported? Partially. His desensitization argument — that repeated exposure to realistic violent simulation reduces the affective response to violence — has stronger empirical support than his direct aggression argument. The claim that video games cause violent behavior is scientifically contested. The claim that they reduce emotional responsiveness to violence is better supported.
What does Grossman say about the morality of killing in war? He maintains a deliberate moral neutrality on whether specific wars are just, while documenting the psychological costs of killing regardless of the war’s justice. He is clear that the costs are not eliminated by justified killing — the moral processing system does not operate on categories of justification. What he argues is that better preparation, including preparation for the psychological aftermath, reduces the cost without eliminating it.
How does the moral injury concept relate to PTSD diagnosis? Moral injury was not a formal clinical diagnostic category when Grossman wrote the book, though he describes the phenomenon clearly. The subsequent clinical literature, particularly the work of Jonathan Shay and Brett Litz, has developed moral injury as a distinct construct with its own assessment measures and treatment protocols. It is increasingly recognized as a significant clinical issue for combat veterans that requires different treatment approaches than fear-based PTSD.
What can civilians take from On Killing? The most directly relevant dimension for non-military readers is the institutional lesson: the way a society prepares people to do morally complex work and the way it supports their reintegration afterward has enormous effects on the psychological outcomes for those people. This applies to emergency services, law enforcement, medical professionals, and anyone else who is regularly exposed to trauma in service of others. The preparation and the support structure are not incidentals — they are among the most powerful determinants of long-term psychological outcomes.
The Historical Pattern — When Societies Failed Their Warriors
Grossman’s historical survey extends beyond Vietnam to examine patterns across multiple societies and multiple eras of warfare, and the consistent finding is that the societies that failed their warriors most catastrophically shared specific characteristics: inadequate preparation for the psychological dimensions of killing, absent or hostile homecoming rituals, and the denial of genuine community support in the reintegration period. Ancient cultures that understood these requirements — that built warrior castes with explicit philosophical and ritual frameworks for killing, dying, and returning — produced dramatically better psychological outcomes than modern societies that send soldiers to war with physical and tactical training but almost no preparation for the moral and psychological dimensions of combat.
The Roman legions had priests. The Japanese samurai had Buddhism and Bushido. Native American warrior cultures had elaborate purification rituals that explicitly acknowledged the moral weight of killing and provided a community-supported path through it. These were not superstitions. They were functionally sophisticated systems for managing exactly the psychological costs Grossman documents — systems that worked because they were honest about the costs rather than denying them.
The modern Western military institution is in some ways less equipped to manage these costs than many ancient warrior cultures, because it operates within a secular framework that provides fewer of the moral and ritual resources that were available in more religious warrior traditions. The military chaplaincy — which Grossman discusses at length — is an attempt to provide some of these resources, with mixed effectiveness. The after-action debrief and the combat stress control unit are other attempts. None of them fully substitute for the comprehensive moral and community framework that the most effective historical warrior cultures built.
The Media Problem — Entertainment vs. Reality
One of the most important and most contested dimensions of Grossman’s work is his analysis of the relationship between entertainment violence and real-world violence. He draws an explicit parallel between the conditioning techniques used to overcome the killing resistance in military training and the conditioning that violent entertainment — particularly interactive video games — provides to civilian populations including children. The argument deserves careful evaluation rather than either dismissal or uncritical acceptance.
The conditioning mechanism Grossman describes is straightforward: the operant conditioning framework of reward (score increase, advancing game levels) paired with realistic simulation of killing produces the same desensitization and behavioral conditioning that military trainers intentionally create. The human-silhouette targets that changed military firing rates are replaced by photorealistic human targets in interactive environments where the player is rewarded for accurate simulated killing.
The empirical question is whether this conditioning produces changes in real-world behavior at magnitudes that are practically significant. The honest answer from the research is: probably, at the margin, for some populations, under some conditions. The effect sizes found in experimental studies of violent media and aggression are modest and the translation from laboratory aggression measures to real-world violence is uncertain. Grossman’s strongest claims about direct causal effects on crime rates are not well-supported by the available evidence. His more modest claim — that extended exposure to realistic violent simulation produces measurable desensitization — is better supported.
What is undisputed is the principle: conditioning works. The mechanisms that can be used to train soldiers to overcome the killing resistance can be engaged by entertainment. The question of whether that engagement at a civilian population level produces outcomes that are practically significant is empirical, contested, and important. Grossman is right to take it seriously even if his specific claims overreach the evidence in places.
Integration with the Resilience Framework
The deepest connection between Grossman’s work and the broader resilience framework this publication develops is the recognition that genuine resilience — the kind that allows human beings to face extreme difficulty without being permanently broken by it — is not purely individual. It is social. It is institutional. It is the product of communities that take seriously their responsibility to prepare people for hard things, support them through hard things, and welcome them back from hard things with the honesty and the ritual and the genuine care that make psychological integration possible.
The individual veteran’s resilience is real and matters. But it’s not the primary variable. The individual placed in a unit with genuine cohesion, trained with honest preparation for the psychological as well as physical dimensions of combat, deployed in an institutional context that provides genuine moral framework for the work, and welcomed home into a community that honors the sacrifice and supports the transition — this individual has outcomes systematically better than the same individual placed in a context with broken versions of any of those elements. The institution and the community are doing most of the work. Individual resilience operates on top of them.
This is the lesson that extends from Grossman’s military context to every context in which human beings are asked to do hard and morally complex things. The firefighter, the ER physician, the prison guard, the social worker — each of them is operating in a context that has a moral and psychological weight that the institutional and community structures either adequately support or inadequately support. The outcomes for the individuals doing the work are substantially determined by the quality of that support. Taking Grossman’s research seriously means taking institutional responsibility seriously — building the preparation, the frameworks, and the genuine community support that make it possible for human beings to do hard things without being destroyed by them.
Related: Born a Crime Summary
The Practical Framework: Applying Killing Summary In Real Life
NOT JUST READING ABOUT IT ›
Join the Wolf Pack — mentorship and accountability for men who are done thinking about change
References
Editorial StandardsCorrectionsMedical DisclaimerAbout Our ContentAffiliate DisclosureSite Map