In a deposition room, the person who wins is almost never the one raising their voice. Trial attorneys who take depositions for a living tend to describe the same pattern: the calmer witness, the one who lets a hostile question sit for a beat before answering, tends to leave the room having given up less ground. The louder side often mistakes the quiet for surrender. It usually is not. It is a person who has already run the math on what escalation buys them and decided the answer is nothing.
That posture, unusual calm in the middle of an argument, gets misread constantly. Family members call it detachment. Partners call it stonewalling. Coworkers call it passive-aggression. The pattern in how people actually learn to handle conflict suggests something less flattering to the person doing the escalating: the calm party has often simply figured out that turning up the volume rarely changes what happens next.
What the escalation research actually shows
Decades of observational work by psychologist John Gottman produced the framework the Gottman Institute now calls the Four Horsemen of the Apocalypse: criticism, contempt, defensiveness, and stonewalling, described by the Institute as communication styles that, according to their research, predict the end of a relationship. Three of the four are escalating moves. The fourth, stonewalling, is the opposite, the listener withdrawing and shutting down. Conflict itself was never the thing that forecast failure. How the conflict was run was.
That distinction matters. It reframes calm not as the absence of caring but as the presence of a learned strategy. Someone who watched a parent’s marriage end in shouting, or watched their own first relationship end the same way, has data. They know what escalation produced. They know what it did not produce.
A 2024 study in Communications Psychology, run by researchers at the University of St Andrews, tested this directly. Couples competed in a lab task where they could blast each other with loud noise, and the researchers found that both partners’ negative emotion predicted reactive aggression. Notably, what they observed was retaliation rather than runaway escalation. The sharper finding was about faces: aggression shifted dramatically depending on whether one partner, both, or neither visibly expressed negative emotion. And when the researchers imposed short breaks, aggression fell, apparently because the delay let arousal drop before anyone could act on it.
Read that from the calm party’s side and it stops looking like withdrawal. Refusing to supply the second half of the exchange is doing, in real time, roughly what the forced break did in the lab.
Calm as a skill, not a temperament
Stanford psychologist James Gross’s process model of emotion regulation distinguishes between suppression, where a person hides what they feel, and cognitive reappraisal, where a person reinterprets the situation to change what they feel. In Gross’s review of the two strategies, suppression fails to reduce the emotional experience itself and carries measurable costs, while reappraisal is generally the more effective of the two.
That is worth pausing on. The person who looks unshakable during a fight is often not born unshakable. They have practiced a specific move: reframing the argument in real time. The insult is a symptom of the other person’s bad day. The accusation is not really about the dishes. The threat is a bluff that has been made before. Reappraisal is not denial. It is a working theory about what is actually going on, applied fast enough to keep the temperature down.
This is also why popular advice encouraging people to vent tends to backfire. A 2024 meta-analytic review by Sophie Kjærvik and Brad Bushman, published in Clinical Psychology Review and covering 154 studies and more than 10,000 participants, found that venting does not reduce anger. What worked was lowering physiological arousal: deep breathing, relaxation, meditation, taking a timeout. Some arousal-raising activities made things worse. Turning the volume up on your own feelings does not drain them. It rehearses them.
Where the folk read goes wrong
The folk read of a very calm person during conflict is that they do not care enough to fight. Sometimes that is true. Often it is not. The behavior looks identical from the outside, which is part of what makes the misread so common.
Consider what operant conditioning would predict. Behavior that produces no reinforcement fades. Behavior that produces the desired outcome strengthens. A person who has spent years watching escalation fail, in family arguments that ended with everyone hurt and nothing decided, in workplace disputes where the louder party got labeled the problem, in custody negotiations where the parent who lost composure lost leverage, has been trained by consequences. Escalation stopped being reinforced. Something quieter took its place.
This is not detachment. It is learning.
A recent piece surveying emotional regulation patterns made a similar point, arguing that going quiet mid-argument is less about being unable to articulate a thought and more about recognising when engaging would only escalate things. The framing there was popular rather than clinical, but the underlying observation lines up with what conflict researchers have been describing for decades.
What this looks like in legal settings
Family court is the clearest laboratory. Anyone who has watched a contested custody hearing knows the pattern. The parent who arrives ready to litigate every past grievance, who interrupts opposing counsel, who visibly reacts to testimony, tends to give judges exactly the impression they should have wanted to avoid. The parent who sits still, answers narrowly, and lets the record speak often walks out with more of what they came for.
Mediators describe the same dynamic in civil disputes. The party who cannot stop escalating usually costs themselves money. Not because judges reward calm as a virtue, but because escalation produces sloppy testimony, contradictions, and openings that opposing counsel is trained to exploit.
Police de-escalation programmes rest on the same insight from the other direction. Departments now spend real money teaching officers to lower the temperature of a volatile encounter rather than match it. The training exists because the instinct to escalate is strong, is expensive when it goes wrong, and does not correct itself without deliberate practice.
None of this is legal advice. Laws vary by state, and how a specific case unfolds depends on facts a general article cannot see. What the courtroom offers here is texture, not instruction: a setting where the cost of escalation is unusually visible and unusually well documented.
The cost that does not show up on the outside
There is a real limit to interpreting calmness as simply learned mastery, and it deserves attention. Staying calm has costs. People who have spent years managing other people’s escalation often describe a kind of quiet exhaustion. The skill is real, and it is also work.
Gottman’s own framework contains the warning. Stonewalling, the fourth horseman, is what withdrawal looks like once it has stopped being tactical: the listener shuts down, turns away, and simply stops responding. Strategic calm and stonewalling can look nearly identical from across the room, and one can decay into the other. A person who has learned that escalation does not work may eventually conclude that engagement does not work either, which is a different and worse outcome.
Workplace research points at a related version of this. A 2016 study of 798 service workers across 55 facilities, published in Frontiers in Psychology, sits within a literature showing that intragroup conflict tracks with worse employee health and worse quality of service. The interesting part was the moderator: a functioning conflict-management climate blunted the damage. Individual composure was not the whole answer. Whether the group had any real mechanism for resolving things mattered too. Calm that comes from strategy is different from calm that comes from having given up, and only one of them is sustainable.
So the picture is not that quiet people have solved conflict. It is that they have solved one part of it, the part where making it louder makes it worse, and they still have to solve everything else.
How the calm party is often reading the room
Attorneys who prep clients for depositions describe drilling one habit above all others: slow down. Wait before answering. Do not fill silence. Do not defend against a question that has not actually been asked. The advice is not about hiding anything. It is about not manufacturing new problems in the middle of an existing one.
The same instinct, applied outside a legal setting, is what a calm partner or coworker is often doing during an argument. They are running a fast triage. Is this the argument, or a proxy for a different one. What outcome is actually available here. What does saying the sharp thing produce that not saying it does not. The calm is not empty. It is full of calculation.
That calculation gets easier with repetition, and it is teachable. A Globe and Mail piece on lessening conflict in adolescence argued that prevention beats reaction, and that the most effective response is sometimes flat neutrality: acknowledge the other person is upset, and come back to it later. The advice was specific to teenagers, but the mechanism is general. Withholding the escalation the other party is trying to draw out changes the shape of the interaction itself.
What the research supports and what it does not
The evidence supports a fairly narrow claim. Hostile conflict patterns predict worse outcomes than de-escalated ones, across relationships and workplaces. Emotion regulation strategies like reappraisal can be learned, and work better than clamping down on the outward signs. Venting does not reliably reduce anger and sometimes worsens it. Imposed pauses reduce aggression in the lab. People who look calm during conflict have often practiced getting there.
The evidence does not support the stronger claim that calm people are always right, or that quiet always beats loud, or that a person who stays composed is automatically the healthier party. Some calm is strategic. Some calm is avoidance. Some calm is exhaustion that has been mistaken for wisdom. The capacity to stay steady under pressure is real and trainable. It is also not free, and pretending otherwise does the people who have built it no favours.
Which is maybe the more useful thing to notice about the person across the table who will not raise their voice. They are not necessarily above the argument. They may just have run this argument enough times, in enough forms, to know where it ends. That knowledge is not detachment. It is a kind of memory.
The louder party often reads that memory as weakness. It rarely is.