A Neighbor's Calm as a Safety Signal express

Original: "Social Discounting Enables Fast and Reliable Collective Escape"
· Zachary P Kilpatrick
arXiv:2608.14965v1 · 2026-08-15 · CC BY 4.0 · 2 min
A school of fish avoids false alarms by trusting the stillness of its neighbors.

A lone fish sees a shadow overhead. Should it flee? If it's just a leaf, it wastes energy and gives away its position to a predator. If it's a heron, hesitation is fatal. That's the eternal trade-off of the solitary individual: quick reactions mean many false alarms, while caution means risking being eaten.

In a school, it's different. A neighbor that calmly keeps nibbling on algae is living proof that there's no danger. Scientists built a model where each fish accumulates ambiguous signals like drops in a glass and darts off when the level overflows. It turns out that a neighbor's flight is strong evidence of a threat, but its stillness is evidence against one. A naive fish reacts only to others' flights and, as the school grows, starts darting at every little thing. A smart fish weighs both: it's as if it discounts the threat while all is quiet around it.

In a large school, even a weak noise can trigger an avalanche of false escapes if everyone reacts only to flight. But if fish can appreciate their neighbors' calm, the avalanche dies before it starts.

This discount has a precise measure. From the school's behavior, one can calculate how much an animal devalues an alarm signal when neighbors are calm. And this number has a limit: if a fish watches too many neighbors, a single discount is no longer enough to suppress false alarms. Observations of wild mollies attacked by birds showed that their discount is much larger than what simple individual accumulation of signals could produce. That's why false alarms in the school don't increase even as the number of fish grows.

Thus, the school turns the solitary individual's weakness, inevitable errors, into strength. A neighbor's calm becomes as important a signal as its flight.

In mollies, false alarms don't become more frequent as the school grows, even though it might seem that the more fish there are, the higher the chance that one of them makes a mistake.

Concepts
false alarm rate prior distribution Price equation bayes factor bayesian evidence random forest bayesian hierarchical model Shelford's law of tolerance lead time probabilistic forecast Gause's competitive exclusion principle density Reinforcement Learning emergent collective learning saddle point statistical uncertainty symbolic regression microbial collective motion saddle node bifurcation sample size