Structural Violence: Johan Galtung's Challenge

In 1969, Norwegian sociologist Johan Galtung asked a question that still reverberates through debates about violence and progress: If people die from preventable causes—poverty, lack of medical care, discriminatory policies—is that violence?

His answer was yes. He called it "structural violence"—harm embedded in social systems rather than inflicted by identifiable actors. No one pulls a trigger. No one commits murder. Yet people die all the same—and die, Galtung argued, as victims of violence.

This distinction matters enormously for the declining violence thesis. Steven Pinker counts direct violence: murders, war deaths, executions. But if structural violence counts too, the picture changes dramatically. The world might be less violent by one definition and more violent by another—simultaneously.

Direct Violence vs. Structural Violence

Traditional definitions of violence involve an actor, an action, and a victim. Someone stabs someone else. An army bombs a city. A mob lynches a person. There's intention, agency, a clear causal chain.

Structural violence has none of these features. Instead, harm results from how society is organized. Consider:

- A child dies from malaria because pharmaceutical companies don't invest in treatments for diseases of the poor. Who killed the child? - A man dies of a treatable condition because he lacks health insurance. His death was preventable—but who is responsible? - A population suffers chronic malnutrition because colonial-era land distributions concentrated farmland among a small elite. The colonizers are dead. The violence continues.

In each case, life expectancy is reduced, human potential is truncated, people die prematurely—but there's no one to charge with homicide. The violence is built into structures: economic systems, legal frameworks, international trade rules, historical inheritances.

Galtung's provocation: if a person dies at 30 when they could have lived to 70, and the shortened life results from remediable social conditions, what's the meaningful difference between that and murder? The person is just as dead. The death was just as preventable. Only the mechanism differs.

The structural violence framework forces uncomfortable questions. When we design systems that predictably kill people—and choose not to redesign them—how different is that from killing directly? If we know a policy will result in deaths and implement it anyway, where does neglect end and violence begin?

The Math of Structural Violence

Galtung proposed a way to quantify structural violence: compare actual life expectancy to potential life expectancy under ideal conditions. The gap between the two represents structural violence.

If a society's actual average lifespan is 60 years but the biologically achievable lifespan under optimal conditions is 75 years, those 15 lost years—multiplied across the population—represent structural violence. People are dying, their lives shortened by forces they can identify but cannot fight in the way they might fight an attacker.

By this measure, structural violence dwarfs direct violence. The World Health Organization estimates that approximately 15 million people die annually from poverty-related causes—lack of clean water, inadequate nutrition, preventable diseases, insufficient medical care. That's equivalent to a major war every year, year after year, with no ceasefire in sight.

The scale is staggering. Direct violence—all the murders, all the wars, all the terrorism—kills perhaps 500,000 people annually in a bad year. Structural violence kills 30 times as many. Every two days, structural violence kills as many people as died in the 9/11 attacks. Yet we don't treat it as an emergency.

Pinker's data shows war deaths declining from peaks of millions per year to current levels of tens of thousands. But structural violence deaths remain in the millions—far exceeding the body counts Pinker tracks. If you include structural violence, human society isn't getting less violent. It's getting differently violent.

Colonial Violence and Its Afterlives

The structural violence framework illuminates patterns that body-count statistics miss.

Consider the British Empire. Direct colonial violence—massacres, executions, wars—was terrible but relatively contained. The Bengal famine of 1943 killed approximately three million people under British rule, more than died in the London Blitz. But this wasn't classified as "violence"—it was administrative failure, economic policy, food distribution decisions. No one was charged with mass murder.

Or consider the slave trade's long shadow. Slavery itself was direct violence. But its structural effects—the impoverishment of African economies, the wealth extraction that funded European industrialization, the intergenerational trauma of descendants—continue to produce differentials in health, wealth, and life expectancy today. The slave traders are long dead. The shortened lives continue.

The structural violence lens reframes "progress" narratives. Europeans became less violent toward each other while continuing structural violence against colonial subjects. The "civilizing process" that reduced murder in London coincided with systems that killed millions in India, Africa, and the Americas—through mechanisms that wouldn't show up in homicide statistics.

This raises an uncomfortable possibility: perhaps violence didn't decline so much as relocate. The peace of the metropole depended on extraction from the periphery. Stability at home required instability abroad. What looks like progress from London looks different from Bengal.

This isn't whataboutism. It's an empirical claim: the scope of your definition determines your conclusions. Narrow the definition to direct violence, and progress is real. Broaden it to structural violence, and the picture is far more ambiguous.

The Poverty of Counting Bodies

Structural violence theorists argue that body counts are inadequate measures of human suffering for several reasons:

Death isn't the only harm. People can suffer terribly without dying. Chronic illness, disability, malnutrition, psychological trauma—these reduce quality of life without producing corpses. A society that keeps people alive but miserable isn't necessarily less violent than one that kills them.

Distribution matters. A society where everyone lives to 70 is different from one where half live to 90 and half die at 50, even if average life expectancy is identical. The latter has structural violence that the former doesn't.

Agency matters. Being murdered and dying from systemic neglect feel different to the victim—but should that phenomenological difference change our moral assessment? Galtung argued no: what matters is the outcome, not the mechanism.

Invisibility matters. Direct violence is visible—it makes news, generates outrage, prompts response. Structural violence is invisible—the daily deaths from poverty don't make headlines. This visibility asymmetry distorts moral attention and policy priorities.

The structural violence critique suggests that Pinker's optimism is partly an artifact of what he chooses to measure. Count bullets and bombs, and things improve. Count shortened lives and preventable suffering, and the picture is darker.

Criticisms of the Structural Violence Concept

The concept has attracted significant pushback:

Definitional stretching. If poverty is violence, what isn't violence? The concept becomes so broad that it loses analytical usefulness. Everything bad becomes "violence," and we can no longer distinguish between different types of harm requiring different responses.

Causal opacity. Direct violence has clear cause and effect. Structural violence involves diffuse, multi-causal chains where responsibility is impossible to assign. This makes the concept difficult to operationalize—who exactly should be held accountable for structural violence?

Category confusion. Murder and poverty are different things. Conflating them obscures the specific moral features of intentional harm. A society that ends murder but not poverty has made real moral progress, even if suffering continues.

Policy paralysis. If everything is violence, priorities become impossible. Should we focus on preventing murders or reducing inequality? The answer matters—resources are limited. Calling both "violence" doesn't help allocate them.

Emotional manipulation. Calling poverty "violence" might be a rhetorical move designed to transfer the moral urgency of violence onto economic questions. This could be politically motivated rather than analytically sound.

Philosophical objections. There's a meaningful moral distinction between doing and allowing, between killing and letting die. Most ethical frameworks distinguish between action and inaction, even when outcomes are similar. Structural violence collapses this distinction in ways that some find philosophically suspect.

The Synthesis: Violence as a Continuum

Perhaps the most productive framing treats violence as a spectrum rather than a binary. At one end: intentional, direct, physical harm. At the other end: systemic conditions that shorten lives and constrain flourishing. In between: various forms of coercion, exploitation, discrimination, and neglect.

By this view: - Direct violence is declining (Pinker's data supports this) - Some structural violence is declining (global poverty rates have fallen dramatically) - Other structural violence persists or worsens (inequality, environmental degradation, climate impacts)

The question isn't whether violence is declining, but which kinds of violence are declining and which are increasing. This nuanced view preserves Pinker's insights about direct violence while acknowledging Galtung's point that counting bodies captures only part of the picture.

It also suggests different interventions for different problems. Reducing murder requires police, courts, and norm change. Reducing structural violence requires economic policy, international development, and system redesign. Pretending these are the same problem guarantees failure at solving either.

What Would Galtung Count as Progress?

Galtung's framework implies a different metric for progress: not just declining death rates but declining gaps between actual and potential human welfare. Progress means:

- Reducing the difference between rich and poor countries' life expectancies - Eliminating preventable diseases that kill millions annually - Ensuring everyone has access to nutrition, healthcare, education, and opportunity - Dismantling systems that advantage some populations at others' expense

By these measures, we've made real but uneven progress. Global poverty has declined substantially. Life expectancy has increased worldwide. Child mortality has plummeted. But enormous disparities remain, and some structural violence—particularly environmental—may be intensifying.

The structural violence lens doesn't negate Pinker's findings. It contextualizes them. Direct violence matters, and its decline is real and important. But it's not the only thing that matters, and treating it as the sole measure of violence obscures forms of harm that kill far more people.

The Takeaway

Johan Galtung's concept of structural violence challenges us to think beyond body counts. If violence is whatever shortens lives and constrains human potential, then poverty, inequality, and systemic neglect are violent—even without perpetrators or weapons.

This doesn't mean Pinker is wrong about direct violence declining. It means that focusing only on direct violence gives an incomplete picture. The world is becoming less violent by some measures and more violent by others—and which measures matter is a moral question, not just an empirical one.

Structural violence reminds us that the absence of war and murder isn't the same as peace. A society where no one is shot but millions die from preventable causes isn't peaceful—it just has different kinds of violence. Understanding this expands our sense of what progress requires.

The debate between Pinker and Galtung isn't really about data. It's about what violence means, what progress means, and what kind of world we're trying to build.

Further Reading

- Galtung, J. (1969). "Violence, Peace, and Peace Research." Journal of Peace Research. - Farmer, P. (2004). "An Anthropology of Structural Violence." Current Anthropology. - Winter, D. D., & Leighton, D. C. (2001). "Structural Violence." In Peace, Conflict, and Violence.

This is Part 4 of the Violence and Its Decline series. Next: "Randall Collins: The Micro-Sociology of Violence"