Someone in your class is wrong. You know exactly why. You have the research, the counterargument assembled. You say nothing.

Part of it is calculation — USC is the kind of place where speech has costs that compound quietly, an institution of higher education where reputation is tedious to build and even more treacherous to maintain. The rest is harder to name: a slow intuition that the argument, even if you won it, would land somewhere and change nothing. That intuition, it turns out, is correct.

The conventional explanation for this silence is that we have a free speech problem. The surveys seem to confirm it: A Foundation for Individual Rights and Expression survey conducted in October 2025, the month following Charlie Kirk’s assassination on the campus of Utah Valley University, found that 45% of students nationally said they were less comfortable expressing their views in class after the shooting; 48% said the same about social media.

A concept now common in First Amendment scholarship gained new campus resonance after Kirk’s assassination: the “assassin’s veto” — the silencing of speakers not by force, but by the credible threat of violent retaliation. In Fall 2024, USC’s own FIRE ranking fell from 109th to 245th out of 257 surveyed institutions following the cancellation of valedictorian Asna Tabassum’s commencement speech.

Universities — including Harvard, Vanderbilt, the University of Chicago and USC — have all launched or funded structured dialogue initiatives in the past three years on the same assumption: What students need is more and better conversation, and if you build the room and hand them the prompts, the conversation will follow.

That assumption is incomplete in a way that makes it almost useless.

The problem on this campus is not that students lack opportunity to speak. It’s that they don’t trust that speaking is safe. Those are not the same diagnosis, and they do not have the same cure.

‘Trust is earned’: What USC actually broke

A report released April 10 by the Yale Committee on Trust in Higher Education documented what has actually happened to American universities over the past decade. Ten years ago, 57% of Americans conveyed a “great deal” or “quite a lot” of confidence in higher education. By 2024, that number had plummeted to 36%.

Among the drivers the committee identified was “an array of issues about what is said and taught on university campuses, including matters of free speech, political bias, and self-censorship.” The report was direct about what repair requires: “Trust is earned by doing what you say you’re going to do.”

That sentence lands differently next to USC’s recent history.

In Spring 2024, Tabassum became USC’s valedictorian — and had her commencement speaking invitation rescinded after her pro-Palestinian stance on social media drew complaints from pro-Israel student groups. USC cited security concerns. Critics called it censorship.

The University then canceled the main commencement ceremony entirely. Among students who FIRE surveyed, 88% said it was clear that USC protected free speech before the encampments; in data after the encampments — after the arrests, the suspensions, the canceled ceremonies — that figure dropped to roughly half.

USC didn’t have a free speech problem. It had a trust problem. It told students their voices mattered, then demonstrated, at the highest-stakes moment available, that they didn’t. The encampments were not a rupture. They were a preview.

The Open Dialogue Project launched last October, following a recommendation from an Academic Senate-Provost task force on academic freedom. President Beong-Soo Kim — who only officially became USC’s 13th president in February — championed it as the institution’s answer.

The project facilitates structured, small-group conversations across political and ideological lines, designed not to debate or persuade, but to build the kind of mutual understanding that precedes productive disagreement. Whether that understanding can survive contact with science is a question research has been answering for years.

This is your brain on politics

The neuroscience explains what USC did not reckon with before 2024. In a 2016 study published in Scientific Reports, Jonas Kaplan, a neuroscientist at USC’s Brain and Creativity Institute and associate professor of psychology, scanned participants’ brains while challenging their political and nonpolitical beliefs with counterevidence. When political beliefs were challenged, the brain lit up in regions associated with personal identity and emotional threat — the amygdala and insular cortex.

“Political beliefs are like religious beliefs,” Kaplan told USC News following the study’s publication, “in the respect that both are part of who you are and important for the social circle to which you belong.”

In an interview with the Daily Trojan in April, Kaplan described what the scans were actually showing — something more precise than stubbornness. Political beliefs, he said, are drawn into what he called the “protected circle of the self” — the part of identity the brain is designed to defend. That protection isn’t only about ego. It’s social.

“If you’re going to change one of those beliefs, you might have to change a social relationship,” Kaplan said. “You might have to change how you talk to and relate to your friends and family. That’s a big ask.”

That is why debate doesn’t work, and why structured dialogue initiatives miss the point. The brain doesn’t open under pressure; it opens under safety. Safety is another word for trust. And when the 2024 commencement controversy forced USC to choose between institutional control and that trust, every student watching learned which one it would prioritize.

Research published in the Peabody Journal of Education in 2021 found that high school students in structured political debate came out more polarized than before. Students placed in a deliberation format — where the goal was reaching consensus rather than winning — moved closer toward agreement.

“Discourse in contemporary American politics,” the researchers concluded, “is more consistent with the objectives of a debate than a deliberation. Participants are more interested in promoting their preexisting ideas than listening to and reflecting on different points of view.”

Universities have had this research for years. They have kept pedestalizing debate.

Dan Schnur, an adjunct professor at the Annenberg School for Communication and Journalism and former political communications expert, has spent decades watching the gap between argument and persuasion play out in both directions.

“We’re really good at listening until we’ve made up our minds,” Schnur said in an April interview with the Daily Trojan. “What we may need to do a better job on is getting students not just to think and argue, but also to listen.”

In his first year of teaching in 1996, Schnur opened his classroom to direct debate on a high-profile California ballot measure: Prop. 209, on racial preferences in university admissions.

“After I wiped the blood off the floor of the classroom,” he said, “I realized that not only had nothing productive happened, but it had been a waste, not just of the students’ time, but of their tuition dollars. They’d spent an hour yelling and screaming at each other.”

He hasn’t done it since. What students need isn’t more opportunity to argue, Schnur said. It’s enough safety to be uncertain.

“If I bring up an opinion with which you disagree, and you tell me you come to a different conclusion but want to know more about what I think, that’s a conversation I want to be part of,” Schnur said. “If I bring up something with which you disagree and you tell me how stupid I am, then maybe I won’t bring up my opinion at all the next time.”

Wändi Bruine de Bruin, provost professor of public policy, psychology and behavioral science at the Price School of Public Policy and director of the USC Schaeffer Institute’s Behavioral Science and Policy Initiative, pointed to research in her field that complicates the premise: The divide students are afraid to cross may be smaller than they think.

“On a lot of topics, Americans think that we’re really polarized, but in reality, we’re not actually that polarized,” Bruine de Bruin said in an April interview with the Daily Trojan.

For example, climate change, she noted, is now far less divisive than most people assume — concern has risen across all political groups as the weather has worsened.

“Maybe what we need to do is come out with some surveys and show where we are polarized and where we are not,” Bruine de Bruin said.

The silence may be partly a response to a phantom phenomenon — a gap that social media has made to feel unbridgeable, and that actual contact might reveal as narrower than advertised. We are afraid of each other in ways the numbers do not support.

“By being on social media, you only perceive more polarization,” Bruine de Bruin said. “While you’re on social media, you’re not really connecting with other people.”

If students are self-censoring in response to a polarization that may not fully exist, then the problem isn’t a shortage of dialogue forums. It’s a shortage of trust.

Confirmation bias, interrupted

This is what Kamy Akhavan learned from 300 million people. As the former CEO of ProCon.org — a nonpartisan reference site presenting both sides of controversial issues without declaring a winner — he watched readers arrive to confirm what they already believed. But, when users found their own views represented more clearly than they could express themselves, their defensiveness dropped.

“When they felt heard,” Akhavan said in an interview with the Daily Trojan, “their trust in the source went up, their credibility perception in their eyes went up and then perhaps, for the very first time, they would look at the other side.”

When he surveyed users on whether they had changed their minds, he expected 2 or 3%. He got 36. The following year, 40.

“The No. 1 lesson I’ve learned in this dialogue space is that listening is the superpower,” Akhavan said.

Feeling heard came first. Trust followed. Openness came after that — and the sequence is not reversible. Most institutional dialogue programs, including USC’s, begin where they want to end: with openness. The assumption is if you give students the room and the prompts, trust will build itself backward.

Akhavan, now associate director of programs for the Open Dialogue Project and managing director of the USC Dornsife Center for the Political Future, is clear-eyed about the limitation.

“It would be an illusion to think you could change all of that within an hour,” Akhavan said.

The project’s bet is that the hour still matters — not because it transforms, but because it expands the circle. This semester, it has run at least 15 programs, drawing students from across USC’s schools and disciplines. In the project’s early days, Akhavan recognized almost everyone who showed up.

“The first one or two, I knew pretty much everyone in the room,” Akhavan said. “Now, I don’t know most people in the room.”

That growth is real. But reach is not the same as trust.

That doesn’t mean Kim’s still-nascent Open Dialogue Project was born futilely. It means the model the program was built on is performing triage on a structural problem. It is, in this sense, an embryonic version of the right answer.

The substrate for conversation

What trust requires, Bruine de Bruin argues, is less structured than any program USC has designed.

“Civic engagement is more than having political discussions,” she said. “It’s about having a community in which people feel connected, safe, happy, valued.”

Polarization worsens, she noted, when people are isolated. Isolation makes it easier to demonize people you don’t know well.

Putting students under a form of spotlight to debate each other highlights differences. Putting them in an event organized around a shared problem — like a community garden or a food bank — builds the substrate on which real conversation can eventually catalyze.

“Being good to people is a value we hold independent of our political views,” Bruine de Bruin said.

Bruine de Bruin pointed to the technique of arguing the opposite of what you believe — asking students to make the case for positions they don’t hold, which makes disagreement less personal and overconfidence harder to sustain.

“What makes people overconfident is if they only think about the reasons why they’re right,” Bruine de Bruin said. “Thinking about why you might be wrong leads to better decisions — and perhaps also a better social climate.”

In his courses, Schnur assigns books across the ideological spectrum — a conservative politician, a progressive journalist — and tells reluctant students they’ll be tested regardless; semester after semester, even the most resistant come to recognize smart people exist on the other side, he said. His instruction on the last day of class: Seek out the smartest version of the argument you most disagree with.

“Any student who feels like they can’t say what they’re thinking because they might be judged or graded harshly,” Schnur said, “is a student who’s being deprived of their full education.”

Kaplan’s research points toward the same conclusion. Influence, he has found, doesn’t travel through argument.

“Influence [and] idea exchange comes better through more intimate settings, less confrontational settings,” Kaplan said. “A conversation with someone you know and respect … the more you understand them to be a person and empathize with them, the more likely you are to understand them and be influenced by them.”

The neural systems that govern identity are the same systems involved in narrative and meaning-making. The brain that closes against an argument opens, sometimes, for a story.

“Conversation is the currency of change,” Akhavan said. “It doesn’t require a Ph.D. degree. It just requires an interest in connecting with another human being.”

But currency requires a functioning economy. Trust is the economy. USC already has the diversity, the proximity and the people. What it has struggled to build is the culture that makes the conversation worth having — and the institutional courage to stop treating that as someone else’s problem.