Sleep on it

We are not at our smartest when we are angry, frustrated or defensive. We are narrower, more reactive and less able to reason cleanly. The best teams learn how to disagree well — but they also know that sometimes the smartest move is not to push through. It is to pause, let the heat come out of the room, and come back after at least one sleep.

There is a point in a difficult conversation where continuing stops being intelligent.

Not because the issue does not matter. Not because disagreement is bad. Not because the people involved are weak.

Because the cognitive state has deteriorated.

When people are angry, frustrated or defensive, the conversation changes. Attention narrows. Threat becomes more salient. Curiosity drops. The urge to win rises. A 2022 systematic review of anger research notes evidence that anger “perceptually and conceptually narrows cognitive scope,” and related work links anger and reactive aggression to weaker response inhibition and executive control. In parallel, broader research on emotion and cognition shows that emotional states shape attention, working memory and decision-making rather than simply sitting alongside them.

That is the part leaders and teams still underestimate. We often talk as though conflict is purely a matter of content — who is right, which decision is correct, what should happen next. In practice, the emotional state of the people in the room changes the quality of the thinking available to them. I would put it simply: when we are sufficiently angry, frustrated or defensive, we are effectively a few IQ points worse. That is not a claim about a literal meeting-room IQ test. It is a practical description of what the research is pointing to: narrower cognition, poorer inhibition and worse judgement under emotional heat.

Healthy disagreement is not the same as pushing through

This distinction matters.

Good teams do need to disagree. They need candour, challenge and the ability to surface tension before it turns into silence or politics. But mature disagreement is not the same thing as endless escalation. There is a point at which trying to force resolution becomes a sign not of courage, but of declining judgement.

Harvard Business Review's 2025 guidance on team conflict is unusually practical here. It explicitly advises taking a break when people are no longer in the right frame of mind, even suggesting coming back tomorrow rather than trying to resolve the issue in a compromised state. The logic is simple: if the conversation matters, the quality of mind you bring to it matters too.

That is why I think timeouts are underrated in serious organisations.

Not dramatic exits. Not punishment. Not avoidance.

A deliberate pause.

A recognition that resolution is unlikely to improve while emotional flooding is still active.

“When we are angry, frustrated or defensive, the room gets dumber.”

Strong emotion narrows the room

One of the reasons conflict becomes so unproductive is that negative emotion tends to compress perspective.

When anger rises, people become more certain of their own interpretation and less interested in alternative frames. They move from “help me understand” to “let me prove why I'm right.” The research language is dry, but the practical effect is familiar: cognition narrows, inhibition weakens and the room gets dumber.

The same literature that documents anger's narrowing effect also points to a broader contrast: positive emotional states tend to broaden attention and thought, while high-arousal negative states can constrict them. That does not mean every difficult conversation needs forced positivity. It means we should stop pretending that emotional state is irrelevant to reasoning quality.

This is also why defensiveness is so destructive. Defensiveness does not merely protect the ego. It changes the objective of the conversation. Once people are defending identity rather than examining reality, learning slows sharply.

“A timeout is not avoidance. It is often the smartest available choice.”

Sleep is not avoidance. It is often the reset

This is where sleep matters.

A lot of people know, intuitively, that things look different in the morning. The research supports that instinct. Reviews on sleep and emotion consistently find that sleep is important for restoring emotional functioning, while sleep loss increases emotional reactivity and weakens emotion regulation. One 2020 review concluded that sleep deprivation makes us more emotionally aroused and more sensitive to stressful stimuli; another highlights decreased functional emotion regulation under sleep problems and deprivation.

There is also direct conflict evidence. A study on couples found that after one night of sleep deprivation, conflict discussions were associated with higher cortisol and less positive emotion than when people were rested. Newer diary evidence points in the same direction: more sleep is associated with less quarreling, while worse sleep is associated with worse conflict dynamics.

So yes, I recommend at least one sleep in the middle when a conflict has become too heated.

Not every disagreement needs that. But some absolutely do.

Because once emotional regulation has deteriorated, a night of sleep is often not procrastination. It is cognitive maintenance.

“If the conversation matters, the quality of mind you bring to it matters too.”

The smartest teams know when to pause

This is part of what maturity looks like.

Immature teams think pausing means weakness. Mature teams understand that pacing is part of good judgement.

They know how to say:

• This matters, and we are not in the best state to solve it well
• Let's pause and come back tomorrow
• I want to continue this when we are both less activated
• We are too defensive right now to make a clean decision

That is not conflict avoidance. It is conflict stewardship.

And the point is not only individual regulation. It is system quality. A team that keeps trying to resolve everything at peak activation will make poorer decisions, create more interpersonal damage and take longer to reach the truth anyway.

A pause only works if you actually come back

There is one important warning here.

A timeout is only smart if it is a timeout, not a disappearance.

The pause should have structure:

• Name that the conversation matters
• Say explicitly that the current state is not helping
• Agree when you will return to it
• Ideally come back with clearer questions, calmer language and a sharper sense of what actually needs resolving

This is where many teams go wrong. They either push through when they should pause, or they pause and never reopen the issue. Neither is good.

The goal is not to avoid tension. It is to return to it in a better state.

The real test is whether the conversation gets better after the break

That is the practical measure.

After the pause, are people more curious? Less certain of their own story? More able to hear? More precise about the actual issue? More able to separate fact from interpretation? More able to say what needs to change without trying to win the whole moral case?

If so, the pause was worth it.

If not, the team may need more than time. It may need stronger norms around feedback, safety, ownership and emotional discipline.

But in many cases, the first fix is simpler than people think: stop trying to finish the conversation in the worst possible state.

A simple rule

Disagree healthily. Do not avoid hard things. Do not confuse pausing with weakness.

But when the room is too angry, too frustrated or too defensive, stop pretending more minutes will produce more intelligence.

Sometimes the smartest move is to step back, let everyone deflate, and come back after at least one sleep.

Because a conversation conducted in a worse cognitive state is not braver.

It is usually just worse.

“A night of sleep is often not procrastination. It is cognitive maintenance.”

“Do not confuse pushing through with healthy disagreement.”