The Harmony Reflex

Why Disagreement Feels Like Danger Before It Feels Like Information

Issue 274, July 23, 2026

A vice president at a financial services client described a moment to me that she had never mentioned to anyone. In a strategy session with eleven colleagues, the group converged on a market assumption she believed was wrong. She had the data. She had spent three weeks on the analysis. When the room turned to her, she heard herself say, “That works for me.” She told me she felt her face flush as the words left her, and that the discomfort had arrived before the decision to agree did. “I didn’t decide to go along,” she said. “I was already going along by the time I noticed I was doing it.”

That sequence is the subject of this issue. Not the structured silence that follows, which we have examined before in this newsletter and on the podcast. We are focusing on the thing that happens roughly a half second earlier, when the mind is still assembling the objection and something older and deeper has already cast its vote.

It deserves a name, because naming it changes what we do about it. I am going to call it the Harmony Reflex: the involuntary physiological response that registers disagreement with others as a threat to belonging to a group, team, or even an organization, before the brain has evaluated whether disagreement with others is the right thing to do. It’s not a decision. It’s not a character trait. It’s not cowardice. It’s a reflex, and like most reflexes, it fires nearly automatically whether or not the situation warrants it. And most often after the reflex fires, we question ourselves, we criticize ourselves, we might even have feelings that we violated ourselves, turned aside what we believe in, we may question our values of right and wrong or even feel stronger feelings that we let ourselves down at the felt pressure from a group of others.

Think about how many times over your career that reflex may have fired while with a group of your peers? Your organization’s leadership team? In a team meeting? It doesn’t even have to happen in person; it can happen as a result of an email chain, a series of text messages and really any situation where more than one person is involved.

What the Brain Does With Rejection

In 2003, three researchers at UCLA put people in a brain scanner and had them play a simple online game of catch. Partway through, the other two players stopped throwing them the ball.

That was the entire experiment. No insult. No criticism. No consequence of any kind. Just being left out of a meaningless game by strangers they would never meet.

The brain responded in the region that processes physical pain. Not a similar region. The same one. It is called the anterior cingulate cortex, and its job is registering how much something hurts. It became more active when people were excluded, and the more active it became, the more distress they reported feeling. Naomi Eisenberger and her colleagues titled the paper “Does Rejection Hurt?” and answered their own question. Social pain runs on the machinery built for bodily injury. Being left out registers as being hurt.

Sit with how small that experiment was. A pointless game. Anonymous partners. Nothing whatsoever at stake. And the brain reached for the same alarm it uses for a broken bone.

Now consider what that alarm does when the room is your leadership team, the stakes are your credibility, and what is on offer is not a withheld ball but the collective judgment that you are the difficult one.

Two years after “Does Rejection Hurt?”, Gregory Berns and colleagues at Emory ran a study that gets closer to what happens in a meeting. Participants judged whether two shapes were the same object rotated, a simple visual task. Before answering, they were shown what a group of peers had supposedly said. Sometimes the group was wrong.

Two findings deserve everyone’s attention.

The first is that when people went along with a wrong group answer, the change showed up in the visual parts of the brain, the machinery that handles seeing. And it happened most strongly when the wrong answer came from other people rather than from a computer. Going along was not only a decision made at the top of the mind about what to say out loud. Something was shifting further and deeper down, in how the object was perceived in the first place.

The second finding is the one I can’t stop thinking about. When people broke from the group and gave the independent answer, the brain’s fear circuitry activated. The amygdala, which handles threat detection, showed increased activity.

Read that again, because the ordinary framing has it backwards. We talk as though agreement is the easy path and disagreement is the thing that requires effort. The biology and inner workings of our brains say the reverse. Agreement is not a reward. Agreement is the absence of an alarm that standing alone sets off. Yikes.

Asch, and the Number Nobody Quotes

Most readers may remember Solomon Asch’s line experiments from 1955. Participants were shown three lines and asked which one matched a reference line. The task was so easy that people working alone got it wrong less than one percent of the time.

Then Asch seated them in a room with several other people, all secretly working for him, who confidently gave the same wrong answer. Roughly three-quarters of participants went along at least once. Across the trials, people conformed about a third of the time.

Those are the figures everyone cites as a view into how easily we seek to conform to others, doubting or ignoring what we know is right (or wrong). The figure that I think matters more is what happened when Asch broke the unanimity. He placed one other person in the room who gave the correct answer.

Conformity collapsed. Not diminished. Collapsed.

One voice. Someone else saying the right thing, and the pressure to conform lost most of its grip.

The Harmony Reflex is not triggered by the presence of disagreement. It’s triggered by the prospect of being the only one in the room, the meeting, or an entire organization. Which means the reflex is not primarily a courage problem to be solved inside individuals. It’s a structural condition of aloneness that leaders either create or dissolve, and most leaders have no idea which one they are doing.

I wrote about one half of this mechanism three years ago, in Issue 114, without recognizing that it had a second half.

That issue examined social proof, Robert Cialdini’s term for the way people copy the behavior of others when they are unsure what to do. I used the example of an audience member who wants to give a standing ovation but waits to see whether anyone else stands first. Cialdini calls the underlying response primitive automaticity, and I described it then as something the subconscious drives rather than something we decide.

What I said at the time was that social proof exerts its strongest pull under uncertainty. When we do not know enough, we look to others to tell us what to do. That is accurate, and it is the version most of us find easy to accept, because it lets the behavior look reasonable. Of course you follow the room when you do not know the answer.

The research in this issue removes that comfort. The vice president in the strategy session was not uncertain. She had the data. She had spent three weeks on the analysis. She knew the group was wrong. And she went along anyway.

Social proof explains why we follow others when we are unsure. The Harmony Reflex explains what happens when we are sure and the room is not. The first is a shortcut. The second is an alarm. Knowing the answer does not switch it off.

The Fear Is Social, Not Punitive

Here is where the research becomes uncomfortable for those of us who have built careers on the language of psychological safety, and I include myself.

In 2003, Frances Milliken, Elizabeth Morrison, and Patricia Hewlin sat down with forty employees across a range of industries and asked a simple question. What do you not say to your boss, and why?

Eighty-five percent said they had been in a situation where they felt unable to raise something they believed was important. Three-quarters of those said the colleagues who knew about the same issue felt the same way.

Then comes the part that matters. When asked why, the most common answer was not fear of being fired. Not demotion. Not retaliation of any kind. The most frequent reason people gave for staying quiet was that they would be seen as a problem, and that it would damage relationships they valued.

Read that again. People are not primarily protecting their jobs. They are protecting their belonging. The thing they won’t risk is not the paycheck. It’s their standing in the room.

That should complicate how we think about the problem, because it means an organization can be entirely safe in the formal sense, with no retaliation, no career consequence, and a leader who genuinely welcomes challenge, and people will still not speak. The threat they are managing isn’t the one the permission or the policy addresses.

John Detert and Amy Edmondson pushed this further in 2011. Across four studies, starting with 190 interviews inside a large multinational organization, they found that people held back even when the thing they were withholding was a constructive suggestion that would have helped everyone, including the boss.

What stopped them was a set of unwritten rules about when speaking up in a hierarchy is risky or out of place. People couldn’t always put those rules into words. Most had never thought about them. And most had not learned them at their current employer.

The rules arrived before the job did. Regular readers will recognize the connection to Structural Silence in Issue 256 and in an episode of The Human Factor Podcast, where I argued that organizations train people not to speak. The points I made across those pieces stand, and I will now add a qualification to both. The organization isn’t writing the curriculum from scratch. It’s reinforcing a curriculum the nervous system brought with it that is learned early and deeply by each one of us.

The Objection to Answer

Now let me argue against myself.

If we are all so conflict-averse and seeking to go along with the crowd at every turn, why is workplace conflict at a record high?

In November 2025, Acas published the largest study of its kind into conflict at work in Great Britain, surveying more than 4,500 working-age adults. Forty-four percent reported experiencing conflict at work in the previous twelve months. That is the highest level ever recorded in a British survey, well above earlier estimates that ranged from twenty-five to thirty-eight percent. Acas puts the annual cost at £28.5 billion.

That doesn’t look like a population avoiding friction. It looks like the opposite.

But look at what people did about it.

Nearly one in five did nothing at all. Among managers who had a conflict with someone who reported to them, more than a third did nothing, a rate more than double their inaction with anyone else. And fifty-seven percent of everyone reporting conflict came away with stress, anxiety, or depression.

The data does not contradict the Harmony Reflex. It locates it. Avoidance doesn’t show up in whether conflict occurs. It shows up in what happens next.

And notice which conflict managers most reliably decline to touch. The one with the person they see every day, whose regard they depend on, where the relationship is closest and the belonging is most exposed. Exactly where Milliken and her colleagues said the fear lives.

There is a second distinction worth making, and the Acas data makes it visible. The most common subjects of conflict were capability and performance, and personal friction between people. Those are conflicts about people.

What organizations get very little of is disagreement about ideas. The assumptions in the deck. The numbers in the model. The strategies nobody wants to be the one to question.

We are not short on friction. We are short on the specific kind that improves decisions. Conflict between people is abundant and mostly costly. Disagreement about the work itself is scarce and mostly valuable. The Harmony Reflex can’t tell them apart, because to the brain’s alarm system, standing alone is standing alone regardless of the subject.

What the Reflex Costs Us

The temptation is to treat all of this as a workplace pleasantness issue in that we all simply want to get along. It’s not. It’s a decision quality issue, and Charlan Nemeth’s work at Berkeley shows why.

In 1986, Nemeth drew a distinction that has held up for forty years.

When a group hears a dissenting view, people think more broadly. They consider the problem from several angles, reexamine their own assumptions, and go looking for information they hadn’t considered. When a group hears only the majority view, thinking narrows. People concentrate on the position in front of them and stop searching.

Here is the counterintuitive part, and it’s the heart of the matter. The dissenting view improves the group’s thinking whether or not the dissenter is right. What changes isn’t the answer. It is the quality of the process the group uses to reach one.

Which means the Harmony Reflex isn’t merely suppressing correct objections. It is removing the friction that makes any group think well, including friction that turns out to be mistaken. We aren’t just losing the good ideas people are afraid to say. We are losing the thinking that even wrong objections would have forced.

In 2001, Nemeth and her colleagues tested the devil’s advocate, the standard organizational workaround in which someone is assigned to argue the other side, against genuine disagreement.

Real dissent produced more original thinking, more honest consideration of the opposing position, and actual changes of individuals’ minds. The devil’s advocate did something else entirely. It left the group more firmly attached to the position it started with.

Read that carefully, because it is worse than the technique simply not working. Assigning someone to argue against the plan appears to strengthen the group’s belief in the plan. We have been running a ritual that inoculates us against the very thing it was designed to provide. A companion study that same year tested three separate versions of the devil’s advocate role and found genuine dissent superior to all of them, concluding that the value lies in authenticity and that role-playing cannot reproduce it.

The devil’s advocate is our attempt to purchase the benefits of dissent without paying its social cost. It doesn’t work, obviously, because everyone in the room knows the difference between a person who disagrees and a person who has been asked to perform the disagreement. They are simply playing a role. The group’s thinking isn’t fooled. Real dissent works because it’s real, which is to say because someone actually stood alone and their brain actually registered it. There is no version of this that is free of cost.

Where This Sits in the Larger Pattern

I have been circling this reflex behavior for years without realizing it until recently. Decision Theater in Issue 254 described organizations mistaking motion for commitment. The Meeting after the Meeting described strategies that lose support in the corridors where the real conversation happens. The Permission Paradox in Issue 269 described people concealing their AI use because disclosure invited judgment. How to Have a Complicated Conversation, all the way back in Issue 47, offered techniques for conversations we now know most people are neurologically primed to exit.

Each of those issues described a different symptom. The Harmony Reflex is the mechanism underneath all of them.

It also completes a circle of thought that has run through this newsletter and through The Truth About Transformation since the beginning. Resistance is signal, not obstacle. I have made that case for years, over and over again even when I knew I wasn’t being heard.

I want to add its mirror. Agreement is also a signal, and we have almost no tools for reading it. When someone resists, we at least know something is there. When eleven people agree in a strategy session, we record alignment and move on, having measured nothing except the reliable operation of a reflex none of us controls.

What Leaders Owe Their People Here

The instinct is to tell people to be braver, have courage, speak up and be honest. The instinct misreads the reflex completely. You can’t exhort your way out of a fear response.

What Asch demonstrated is that the reflex responds to a structural condition, not a moral one, and structural conditions are exactly what leaders control. Someone has to not be alone. That is the whole finding. Whether that means assigning genuine positions rather than devil’s advocacy, gathering views before the room converges rather than after, or simply noticing when a decision has been reached with suspicious ease and refusing to bank it, the design principle is the same. The question is never whether your people are brave enough. The question is whether your meetings systematically manufacture the condition of aloneness that the brain is built to avoid.

It also requires leaders to separate the two things the reflex can’t separate. Disagreement about an assumption isn’t conflict with a person. Most organizations have never made that distinction explicit, and so people, sensibly, treat every objection as a relationship risk. If you want the first kind and not the second, you will have to say so, repeatedly, and then demonstrate it in the moment someone tests whether you meant it.

And it requires a harder acknowledgment. If the fear people are managing is social rather than punitive, then the absence of retaliation in your organization proves nothing. You can have no consequences, an open door, and a genuine appetite for challenge, and still preside over rooms where nobody says the thing. The reflex does not read your policies. It reads the room.

The Question Underneath

The vice president who told me “that works for me” was not weak although she felt like she was. She was a competent professional whose body reached a conclusion before her judgment could, in a room that gave her no reason to think anyone else saw what she saw.

Which raises the question I would leave with any leader reading this. When your last significant decision met no resistance, what did you remember? Almost certainly alignment. Almost certainly you moved on.

But the smoothness of that meeting was data, and it was telling you something. It may have been telling you the decision was sound. It may have been telling you that eleven people felt their faces flush and said it worked for them.

You cannot tell those apart from the outside. That’s the point. And it means the most confident moments in your organization are the ones you understand least.

Related Reading

Directly Referenced in This Issue

Structural Silence: Why Organizations Train People Not to Speak (Issue 256)

The system-level counterpart to this issue. Structural Silence argues that organizations teach people that quiet is safer than truth. The Harmony Reflex argues that the nervous system arrives already believing it, and that the organization is reinforcing a lesson rather than originating one.

The Permission Paradox: Why Your People Are Hiding the Very Things You Say You Want (Issue 269)

Concealment driven by fear of judgment rather than fear of punishment. The Milliken research cited here explains why that distinction holds.

Decision Theater: When People and Organizations Mistake Motion for Commitment (Issue 254)

Where this issue examines why disagreement does not surface, Decision Theater examines what organizations do with the frictionless consensus that results.

Social Proof (Issue 114)

The other half of this mechanism. That issue explained why we copy others when we are uncertain. This one explains what happens when we are certain and break from the room anyway.

How to Have a Complicated Conversation (Issue 47)

The practical companion, written before I had a name for what makes those conversations physiologically difficult in the first place.

The Truth About Transformation: Revised and Expanded

Resistance is signal, not obstacle. This issue offers the mirror argument: agreement is signal too, and we have almost no instrumentation for reading it.

Human Factor Podcast

Season 2, Episode 019: Structural Silence, Why Organizations Train People Not To Speak

Season 2, Episode 018: The Organizational Immune System, When Culture Attacks What It Doesn’t Recognize

Research Sources

1. Eisenberger, N. I., Lieberman, M. D., & Williams, K. D. (2003). Does Rejection Hurt? An fMRI Study of Social Exclusion. Science, 302(5643), 290–292.
2. Berns, G. S., Chappelow, J., Zink, C. F., Pagnoni, G., Martin-Skurski, M. E., & Richards, J. (2005). Neurobiological Correlates of Social Conformity and Independence During Mental Rotation. Biological Psychiatry, 58(3), 245–253.
3. Asch, S. E. (1955). Opinions and Social Pressure. Scientific American, 193(5), 31–35.
4. Milliken, F. J., Morrison, E. W., & Hewlin, P. F. (2003). An Exploratory Study of Employee Silence: Issues that Employees Don’t Communicate Upward and Why. Journal of Management Studies, 40(6), 1453–1476.
5. Detert, J. R., & Edmondson, A. C. (2011). Implicit Voice Theories: Taken-for-Granted Rules of Self-Censorship at Work. Academy of Management Journal, 54(3), 461–488.
6. Nemeth, C. J. (1986). Differential Contributions of Majority and Minority Influence. Psychological Review, 93(1), 23–32.
7. Nemeth, C. J., Connell, J. B., Rogers, J. D., & Brown, K. S. (2001). Improving Decision Making by Means of Dissent. Journal of Applied Social Psychology, 31(1), 48–58.
8. Nemeth, C., Brown, K., & Rogers, J. (2001). Devil’s advocate versus authentic dissent: stimulating quantity and quality. European Journal of Social Psychology, 31(6), 707–720.
9. Acas (2025). How prevalent is individual conflict at work in Great Britain in 2025? Published 20 November 2025. Survey conducted by the National Centre for Social Research, August to September 2025, 4,558 respondents.