Punishment is one of the stranger things social animals do. It costs time, energy, and sometimes safety — yet mammals do it anyway. Humans are especially enthusiastic punishers, but we are not the only ones. Rats retaliate, chimpanzees take revenge, and under some conditions rats can even intervene on behalf of a stranger.

So how similar are humans and rats when it comes to punishment?

The short answer: both punish, but humans punish more broadly and seem to enjoy it more.

What counts as punishment?

In behavioral science, punishment usually means imposing a cost on another individual after a perceived transgression. There are two useful categories:

  • Second-party punishment: you punish someone who harmed you.
  • Third-party punishment: you punish someone who harmed someone else, even though you were not directly affected.

Many mammals do the first. Humans are unusual because we do the second so readily — and often at a personal cost.

Humans: punishment as fairness enforcement

Humans punish for at least two reasons:

  1. Retribution: to “even the score.”
  2. Deterrence: to discourage future wrongdoing.

But human punishment goes further. We punish strangers who violate norms. We support institutions that punish on our behalf. And we often willingly pay a cost to punish someone who behaved unfairly, even when we gain nothing personally. This is called altruistic punishment, and it is thought to help sustain cooperation in large groups.

There is also a neural reward component. Neuroimaging studies have found that punishing a cheater can activate reward-related brain regions, including the caudate nucleus. In other words, revenge can feel satisfying. That does not mean punishment is only pleasure-seeking, but the anticipation of satisfaction can motivate punitive choices.

Rats: retaliation is common, third-party punishment is conditional

Rats clearly punish in the second-party sense. They retaliate against aggressors, and their retaliatory behavior can depend on their history with a specific partner — suggesting they track social scorekeeping.

Third-party punishment is trickier. A 2024 study found that rats did not usually intervene when they witnessed an aggressive mouse attacking a victim. But rats that had been extensively handled with affection by human caretakers did intervene — gently contacting the aggressive mouse and reducing aggression toward the victim, even when the aggressive mouse fought back.

That suggests third-party punishment-like behavior in rats is not necessarily innate. It may be an acquired behavior shaped by early social experience and nurturing conditions.

Do rats “like” to punish?

This is where the comparison gets cautious. Rats learn from punishment. Dopamine circuits in the ventral tegmental area and prefrontal cortex help encode action–punishment relationships. But there is no strong evidence that rats derive hedonic pleasure from punishing others in the way humans can.

So the safer claim is:

Rats punish, retaliate, and can intervene under some conditions — but there is little evidence they have a human-like appetite for punishment.

Humans vs. Rats: a quick comparison

FeatureHumansRats
Second-party retaliationCommonCommon
Third-party punishmentRobust and widespreadConditional; observed after affectionate handling
Personal cost to punishOften acceptedCan occur in intervention experiments
Neural reward for punishmentEvidence for reward-circuit activationDopamine involved in punishment learning, not clearly “punishment pleasure”
Main functionNorm enforcement and cooperationSelf-defense, social learning, conditional helping

The takeaway

Punishment is not uniquely human. Rats retaliate, and under nurturing conditions they can intervene against aggressors. But humans are unusual in how readily we punish third parties, how much we care about fairness norms, and how often punishment recruits the brain’s reward system.

In other words: rats punish, but humans punish — and sometimes enjoy it.