Yesterday in American Psychologist, a team of neuroscientists published the results of a study they’d conducted in rural Rwanda. They’d traveled the country with portable EEG equipment and laptops, strapping electrodes to the heads of people who had lived through the 1994 genocide against the Tutsi, and they’d asked those people to do something simple: follow an instruction to take money from another person, or refuse.
The participants fell into three groups—former perpetrators, bystanders, and rescuers who had defied orders to save lives—and the researchers wanted to know whether the brains of these three groups processed moral decisions differently. What Emilie Caspar and her team at Ghent University found is one of those results that confirms something you already suspected and makes it worse.
Because the bystander brain and the perpetrator brain are indistinguishable.
Under EEG, former perpetrators and bystanders showed no stable neural differences. Not in emotional processing, not in decision-making patterns, not in how they responded to another person’s distress. The people who looked away and the people who committed atrocities are, at the level of measurable brain activity, the same. The situational pressures—propaganda, dehumanization, fear, tribal loyalty—did the sorting. Not character. Not some essential darkness in one group that the other lacked.
The only group that was neurologically distinct was the rescuers. Their brains showed stronger emotional responses to others’ suffering—a marker called the Late Positive Potential that indicates heightened empathy processing. They reflected longer before acting. They didn’t look away. They also cited a wider range of motivations for disobedience, including family upbringing and historical awareness. And more often than either perpetrators or bystanders, they reported having grown up with role models who had themselves engaged in rescue. Someone had shown them what resistance looked like before they needed to do it.
Hannah Arendt named the phenomenon in 1961. Sixty-five years later Emilie Caspar discovered the mechanism.
Last month I published an essay about a friend who called my concerns with creeping authoritarianism in America “boring and lacking substance.” He’s a good guy. He’s not a fascist. He’d help his neighbors. But he kept dismissing each individual argument, kept making one small accommodation at a time, because engaging with what I was showing him would cost him something—his sense of himself, his community, his peace of mind. So he refused to engage. He said I was a nutjob. He told me to get my shit together.
He moved on.
I described that as the banality of evil, Arendt’s insight that atrocity doesn’t require villains, only people who’ve decided that looking away is easier than looking. I didn’t have the neuroscience then. Now I do, and it’s much worse than I thought, because the Caspar study demonstrates why knowing the arguments and the history doesn’t make a difference if the person hasn’t been shown what refusal looks like and lacks the neural capacity to actually feel someone else’s distress hard enough that compliance becomes impossible.
My friend’s read the quotes. He knows the litanies. He can probably recite Niemöller from memory. And none of it matters, because the study’s central finding is that people like him—the ones who look away, who find the arguments boring and lacking substance, who move on—are neurologically indistinguishable from the ones who committed the atrocities. The bystander brain and the perpetrator brain are the same brain. Knowledge without behavioral role modeling doesn’t produce resistance. It produces people who know what they should do and don’t do it.
The Republican response to Operation Metro Surge is this study playing out in real time, with the three groups labeled and sorted for you.
The loyalists are the perpetrators—not in the sense that Ron Johnson personally carried out raids, but in the sense that they actively enabled and defended the project. Johnson told Minnesota’s attorney general he should “feel damn guilty” for encouraging people to protest. Stephen Miller called a peaceful, lawful gun owner murdered by federal agents a “domestic terrorist.” The White House labeled him a “would-be assassin” in direct contradiction of video evidence.
These aren’t bystanders. They’re participants.
The uncomfortable middle is the bystander group, and the Caspar study predicts their behavior with eerie precision. Ted Cruz urged the administration to be “more measured.” Greg Abbott said the White House should “recalibrate.” Lisa Murkowski called the killing “shocking” and asked serious questions about training. Bill Cassidy said the credibility of ICE was at stake. Every one of them expressed concern about tactics while affirming the goal. They called for investigations they had no power or apparent intention to compel. They used words designed to register discomfort without requiring action—“recalibrate,” “more structured,” “serious questions”—the linguistic equivalent of the bystander brain’s emotional flicker that never reaches the threshold of disobedience.
Speaker Mike Johnson never issued a statement at all. Often the bystander response is simply silence, and silence is the purest form of it.
And then there’s Chris Madel, the Republican gubernatorial candidate who ended his campaign and left the party. “I cannot support the national Republicans’ stated retribution on the citizens of our state,” he said, “nor can I count myself a member of a party that would do so.” He added: “I have to look my daughters in the eye and tell them, ‘I believe I did what was right.’”
One man. Out of the entire GOP. The rescuer data point.
The Caspar study found that rescuers were distinguished not by superior knowledge or political sophistication but by the depth of their emotional response to others’ suffering and by the models they’d been given in childhood for what resistance looks like. Madel didn’t cite policy analysis. He cited his daughters. The cost was real, and he paid it anyway, because the alternative was a version of himself he couldn’t live with.
The base, meanwhile, holds. Three-quarters of Republicans still approve of ICE. Seventy-seven percent say the agency is making Americans safer. Seventy-five percent see protesters as acting unlawfully. The uncomfortable middle expressed their discomfort and went quiet. No Republican senator has called for the operation to end. No Republican has introduced legislation to constrain ICE. The DHS funding bill passed with Republican votes.
This isn’t a knowledge problem. These aren’t people who haven’t heard the arguments. They’ve heard them and found them boring and lacking substance, or heard them and decided the suffering is deserved—which is the dehumanization variable that Caspar identifies as a central driver of mass violence. Not a neural deficit. A choice about whose pain counts.
But it’s not everyone. A CNN poll from last week found Trump’s approval among independents has dropped to 26%—a fifteen-point decline over the past year and the lowest of either term, and sixty-one percent say his policies are moving the country in the wrong direction. Sixty-five percent of Americans say ICE has gone too far, an eleven-point increase since last summer. Even among Republicans, strong approval has dipped below 50% for the first time in this term, and younger Republicans are twice as likely as older ones to say he hasn’t had the right priorities. The base holds in aggregate, but the edges are fraying.
So, something’s shifting in the middle who’ve been watching what’s happening—and I suspect it’ll shift even farther in response to the strikes launched against Iran this morning without congressional authorization. People are feeling the weight of things in a way that goes beyond knowledge into something the Caspar study would recognize as experience. Not the experience of living through genocide, thank God, but the experience of watching your country do things you didn’t think it would do, and discovering that the discomfort isn’t abstract anymore. It’s in your neighborhood. It’s on your screen. It’s costing you something to look away.
While the research is troubling, it’s also hopeful, because it shows us resistance can be learned. Not from textbooks, not from litanies, not from knowing the history—but from role models who’ve demonstrated what it looks like to refuse. The rescuers didn’t know more than the bystanders. They’d been shown more. Someone in their life had drawn a line and held it, and when the moment came, they had a template for what refusal looks like.
That’s what I’m trying to do for my children. That’s the work now. Not more arguments. Not more evidence. Not more history lessons for people who’ll find them boring and lacking substance. The work is modeling refusal so that the people around us—our children, our communities, the ones who are watching and haven’t yet decided what they are—have something to recognize when the moment comes for them.
Chris Madel looked at his daughters and decided what he was willing to be. I looked at mine back in September and made the same choice. Not because we knew more than anyone else, or had better arguments, or had studied more history, but because we had the right role models and our empathy response fired hard enough to trigger non-compliance. That’s not a policy position. It’s the same neural event Caspar’s EEG measured in the brains of Rwandan rescuers thirty years after they’d risked everything while others looked away.
And according to Caspar’s research, it’s the only thing that’s ever made the difference.
Discover more from The Annex
Subscribe to get the latest posts sent to your email.
“A choice about whose pain counts.” Yep, who counts as a person (let alone actually being One of Us) is huge.
“knowing the arguments and the history doesn’t make a difference if the person hasn’t been shown what refusal looks like and lacks the neural capacity to actually feel someone else’s distress hard enough that compliance becomes impossible.” The point at which it becomes real instead of merely today’s tragic news segment before the segment on the fluffiest scrambled eggs (remember to add milk and whisk, whisk, whisk!).
Milk is the key, and a good whisking. It’s all in the wrist, really.