Male mice that watched other mice fight were better prepared to defend themselves against stronger, more aggressive opponents than mice that did not watch fights, according to a paper published in Nature.
Researchers randomly assigned male mice to groups of fighters, spectators or novices. Over eight days, the mice completed three 10-minute training sessions each day. Novices played alone with stuffed mice, while spectators watched pairs of fighter mice from behind a transparent barrier, allowing them to see, hear and smell the fights.
On the ninth day, each mouse faced a stronger opponent in three 4-minute rounds. Researchers filmed the encounters and tracked activity across 12 interconnected brain regions, using brain cells engineered to glow green when excitatory cells activated and red when inhibitory cells activated.
The researchers used a resident-intruder assay to prompt fighting. “Mice are naturally territorial,” co-author Jorge Iravedra-Garcia of Princeton University told Ars. “A traditional behavioral assay that experimenters use is called resident intruder. You just drop mice that are bred to be slightly weaker, and a little bit more docile, into the home cage of strains we know are more on the aggressive side. Those strains are typically very good at just naturally evoking that behavior.”
The encounters were deliberately unequal, with the larger, more experienced mouse fighting in its home cage. “We tip the scales in favor of having a clear winner and a clear loser by having the clear winner be at home, be bigger, and be more experienced,” co-author Annegret Falkner, also of Princeton, told Ars. “We really wanted to show that these changes are adaptive in some way. Animals don’t know what the outcome of a fight is before they enter it. So we set up a situation where the outcome is a little bit more uncertain for the mouse.”
Spectator mice performed better than novices during the fights. Both spectators and fighters spent less time with their backs to opponents and attacked more quickly than novices. Spectators also used moves similar to those of fighters, although practical experience still gave the fighters an advantage.
The brain activity of spectator mice also became more like that of fighters across all 12 regions. During training, spectators’ inhibitory cells resembled those of fighters, while their excitatory cells were closer to those of novices. On fight night, both types of cells more closely resembled those of fighters.
“When you’re watching a fight, you’re paying attention to many things at once,” said Iravedra-Garcia. “You’re not just simulating that aggressive behavior in your head, you’re also trying to gauge who’s dangerous, who’s not, who’s actually involved in the fight versus ‘Do I know them? Are they novel to me?’ Witnessing aggression leaves a circuit trace that changes how an animal behaves in its next conflict, and that trace looks similar to the one left by fighting itself.”
“They can watch males walking around all they want and none of this happens,” said Falkner. “There’s something specifically salient and arousing about attack that makes them both want to attend to it and also drives this specific plasticity mechanism. That’s one of the open questions: why is attack so sticky? Probably because it creates these adaptive behavioral changes.”
Previous research has reported similar findings, including studies involving macaques. Falkner said the new study identifies a possible mechanism behind the learned behavior. “There’s this broad decrease in regions that are known to provide inhibitory control for some of these aggression control networks,” she said. “This one piece of the aggression network is really just the tip of the iceberg in terms of how global networks for aggression control respond during these observed activities. I think these subcortical subnetworks are really recurrent. It’s a tiny piece of a larger network.”
Whether the findings in mice apply to people remains an open question. “Although it’s difficult to draw a one-to-one parallel, these circuits are extremely conserved, especially these brain regions that seem to show this robust change in inhibitory tone,” said Falkner. “We think about violence as being maladaptive in human culture, but if you are a mouse or an animal who’s around a lot of aggression, it’s actually highly adaptive for you to be ready to fight at any particular time.” The same could be true of humans, she said.
The researchers plan to study whether female mice that watch males fight change their mating preferences. They also want to test whether they can artificially trigger the brain activity changes and related behavior. “It doesn’t seem that farfetched to me,” said Falkner. “I think we know the state that the circuit needs to be in to execute this particular behavioral strategy. So whatever pattern of stimulation we can use to kick this plasticity mechanism into gear seems reasonable.”
DOI: Nature, 2026. 10.1038/s41586-026-11217-9.
