The Phantom Extremist. How Perception Fuels Political Violence

[This article] should also redirect our attention from the question "which side is more violent?" to the one that actually matters: "why do we believe our neighbors want us dead, and what are we willing to do based on that false belief?" - Concordia Discors Magazine

In September 2025, Daniel Byman and Riley McCabe of the Center for Strategic and International Studies published an analysis of 750 terrorist attacks and plots in the United States spanning three decades. Their headline finding detonated across the media landscape: for the first time in more than thirty years, left-wing terrorist incidents outnumbered those from the far right. The Atlantic ran the story under the title "Left-Wing Terrorism Is on the Rise." Conservatives treated the data as vindication. Progressives attacked the methodology.

Within weeks, Charlie Kirk, the thirty-one-year-old founder of Turning Point USA, was assassinated by a lone gunman at Utah Valley University, and the study's findings seemed to materialize in blood. Five months later, across the Atlantic, a twenty-three-year-old Catholic activist named Quentin Deranque was beaten to death by masked assailants outside Sciences Po in Lyon, France, during clashes surrounding a conference by a hard-left member of the European Parliament. In both cases, before the investigations had concluded, the dead were already conscripted into partisan narratives.

The CSIS data merit serious attention. Left-wing violence in the United States has risen measurably since 2016, driven by a combination of anti-government extremism and partisan hostility toward the Trump administration and its institutional instruments. Of forty-one left-wing incidents catalogued since 2016, seventeen were motivated by anti-government sentiment, eleven by partisan extremism. The trend is real. Yet the data that generated the most alarming headlines rested on five incidents in six months, a sample so slender that Amy Cooter of the Institute for Countering Digital Extremism observed it could not support robust inference. The study's own authors acknowledged that the more striking finding was the precipitous decline in right-wing attacks in 2025.

The deeper problem, however, is not which side is more violent in any given six-month window. It is that the categories themselves have become instruments of political combat rather than tools of analysis.

The deeper problem, however, is not which side is more violent in any given six-month window. It is that the categories themselves have become instruments of political combat rather than tools of analysis.

Consider what the CSIS study meant by "right-wing terrorism": the label encompassed white-supremacist massacres, anti-government militia plots, incel violence, QAnon conspiracies, anti-abortion attacks, and partisan extremism directed at left-wing targets. "Left-wing terrorism" swept together anti-capitalist sabotage, environmental militancy, anarchism, pro-LGBTQ+ violence, and attacks on government institutions.

These are not spectra with coherent internal logic; they are, as former FBI Director Christopher Wray once said of modern extremists' beliefs, a "salad bar of ideologies." A white-supremacist gunman and a libertarian who threatens a government official share almost no ideological DNA, yet both may appear in the same dataset under the same heading. The mirror image holds on the left: a committed Marxist-Leninist and a distraught young man who murders a conservative commentator after months of algorithmic radicalization inhabit entirely different moral and psychological universes.

Critics of the CSIS study, writing in Just Security, pressed this point further. They noted that a seventeen-year-old arrested for plotting to assassinate President Trump was excluded from the right-wing tally because his motivations were rooted in neo-Nazi accelerationism rather than conventional right-wing politics, even though most researchers would place accelerationist ideology squarely within the far-right constellation. More than thirty antisemitic plots during the same period defied left-right classification altogether, with perpetrators motivated by causes ranging from pro-Palestinian ideas to neo-Nazism.

The Tesla arson attacks were excluded as economic vandalism; the killing of two Israeli embassy staffers was classified as "ethnonationalist." Each of these coding decisions is somewhat defensible in isolation; taken together, they reveal how profoundly the attempt to sort political violence into binary categories distorts the phenomenon it claims to measure.

This matters because the labels do not remain in academic papers. They migrate into political rhetoric, where they cease to describe and begin to accuse. Using the left-right binary as a weapon, attributing the violence of individual extremists to an entire political formation, is always problematic. The left is not innocent of this maneuver, either.

Progressives who describe all conservative movements as incipiently fascist, or who treat every act of right-wing violence as evidence of a unified authoritarian conspiracy, are engaged in the identical distortion from the opposite direction. Both sides collapse a differentiated political reality into a Manichaean fable that licenses precisely the escalation it claims to oppose.

Both sides collapse a differentiated political reality into a Manichaean fable that licenses precisely the escalation it claims to oppose.

And here we arrive at what may be the most consequential finding buried in the CSIS report, a finding that received a fraction of the attention lavished on the headline numbers. The study noted that fewer than four percent of Americans express support for partisan violence such as assault, arson, or murder. Yet Democrats believe that 45.5 percent of Republicans endorse partisan murder, and Republicans believe that 42 percent of Democrats do. The gap between reality and perception is not a footnote; it is arguably the central engine of the crisis.

The study noted that fewer than four percent of Americans express support for partisan violence such as assault, arson, or murder. Yet Democrats believe that 45.5 percent of Republicans endorse partisan murder, and Republicans believe that 42 percent of Democrats do. The gap between reality and perception is not a footnote; it is arguably the central engine of the crisis.

A 2022 study published in the Proceedings of the National Academy of Sciences by Mernyk, Pink, Druckman, and Willer confirmed this pattern with experimental precision. Across four nationally representative studies, they found that both Democrats and Republicans overestimated their opponents' support for political violence by between 245 and 442 percent. The researchers then tested whether correcting these misperceptions could reduce support for violence. The results were remarkable: a brief informational intervention, simply showing participants the actual (low) levels of their opponents' support for violence, reduced subjects' own support for violence by 34 percent and their willingness to engage in violence by 44 percent. The effect persisted a month later (Mernyk, Pink, Druckman, and Willer, "Correcting Inaccurate Metaperceptions Reduces Americans' Support for Partisan Violence," PNAS, 2022).

The implication is profound: a significant share of the appetite for political violence in America is driven not by genuine ideological commitment but by a false picture of the "enemy", a phantom composed of algorithmic distortion, cable news hysteria, and the ancient human tendency toward what social psychologists call motive attribution asymmetry, the belief that our side acts from love while the other side acts from hate.

The implication is profound: a significant share of the appetite for political violence in America is driven not by genuine ideological commitment but by a false picture of the "enemy", a phantom composed of algorithmic distortion, cable news hysteria, and the ancient human tendency toward what social psychologists call motive attribution asymmetry, the belief that our side acts from love while the other side acts from hate.

This is where the philosophical resources of the tradition that the Concordia Discors Project exists to defend become indispensable. Simone Weil, writing during the catastrophe of the Second World War, diagnosed a condition she called deracinement, uprootedness, as the characteristic spiritual pathology of modern life. In The Need for Roots (1943), she argued that human beings deprived of living connections to community, place, and the past become susceptible to ideological capture, not because they are wicked but because they are empty. "Whoever is uprooted himself uproots others," she wrote. "Whoever is rooted himself doesn't uproot others" (Simone Weil, The Need for Roots, 1943). The observation is a diagnosis of the soil in which violence germinates. Tyler Robinson, Kirk's alleged killer, was a twenty-two-year-old Wordle enthusiast and Magic: The Gathering player with no history of political organizing. The masked men who killed Quentin Deranque in Lyon were street fighters whose political identity had replaced every other form of belonging.

These are profoundly different cases, but both emerge from environments in which political allegiance has become the primary locus of meaning, displacing the thicker, more variegated forms of identity, religious, familial, vocational, communal, that once absorbed and domesticated political passion.

These are profoundly different cases, but both emerge from environments in which political allegiance has become the primary locus of meaning, displacing the thicker, more variegated forms of identity, religious, familial, vocational, communal, that once absorbed and domesticated political passion.

Hannah Arendt arrived at a convergent insight from a different angle. In The Origins of Totalitarianism (1951), she argued that the loneliness of atomized, "superfluous" individuals creates the preconditions for ideological extremism, because ideology offers the uprooted a "fictitious home" and a coherent, if false, account of the world. The sorting mechanisms of contemporary life, algorithmic echo chambers, epistemic bubbles, the collapse of cross-cutting associations, are not totalitarian in Arendt's precise sense, but they reproduce the psychological conditions she described: a world in which citizens no longer encounter one another as complex persons but only as representatives of hostile abstractions.

The sorting mechanisms of contemporary life, algorithmic echo chambers, epistemic bubbles, the collapse of cross-cutting associations, are not totalitarian in Arendt's precise sense, but they reproduce the psychological conditions she described: a world in which citizens no longer encounter one another as complex persons but only as representatives of hostile abstractions.

The way out of this impasse cannot be merely procedural. Better data will help; more honest classification of political violence is necessary and overdue. But the Mernyk study points toward something more fundamental: the cycle of violence can be interrupted at the level of perception. When citizens are shown that their opponents are not the monsters they imagine, their own appetite for aggression diminishes. This finding should humble every commentator, politician, and media platform that profits from inflating the perception gap. It should also redirect our attention from the question "which side is more violent?" to the question that actually matters: "why do we believe our neighbors want us dead, and what are we willing to do based on that false belief?"

It should also redirect our attention from the question "which side is more violent?" to the question that actually matters: "why do we believe our neighbors want us dead, and what are we willing to do based on that false belief?"

We cannot pretend that this question admits of easy resolution. But we can insist that the question be posed honestly. The deaths of Charlie Kirk and Quentin Deranque are real. The grief of their families is real. The political violence that claimed them requires investigation, prosecution, and condemnation without partisan qualification. What is not real, or at least not nearly as real as we have been led to believe, is the picture each side carries of the other: the imagined army of forty percent who would kill for their politics. That bipartisan phantom is the true extremist. Dissolving it may be the most urgent act of civic repair available to us. ◳