Unlike Voters, Fish Make Better Group Decisions

This article was originally posted on RealClearScience.

The entire idea of democracy rests upon the notion that large groups of people will, more often than not, make prudent decisions. In theory, all the stupid voters will cancel each other out, and society’s collective intelligence will result in the best candidates getting elected. However, American voters, particularly since 1992, have almost single-handedly challenged the idea of the “wisdom of crowds.”

Still, collective intelligence exists, despite Americans’ vigorous attempts to disprove it. Large groups of animals are known to make synchronized, life-or-death decisions with great rapidity. Think of a school of fish evading a predator, or a flock of birds finding refuge in a storm. How are those coordinated decisions made? Scientists are just now beginning to unravel how such complex behavior works.

New work from a group led by Angelo Bisazza sheds more light on this decision-making process. But instead of studying large groups (10 or more individuals), which is customary for this type of research, Bisazza’s team looked at how pairs (called “dyads”) of fish made decisions. Specifically, they wanted to determine if a pair of fish was better at making basic numerical calculations than a single fish alone.

Fish can’t count, but they do have a vague sense of relative magnitude known as “numerosity.” For instance, if school of fish “A” is twice as large as school of fish “B,” a fish will be likelier to join the larger group, presumably because it is safer for a fish to swim around with a larger group than a smaller one. The researchers took advantage of this natural instinct in their experiments. (See figure.)

As shown in figure A, Bisazza’s team placed either a single or a dyad of fish into a middle tank. On each side was another tank filled with either 4 or 6 fish. The test was straightforward: Which tank would the fish swim toward? Figure B depicts the result: Single fish were indecisive, spending roughly equal amounts of time with both the group of 4 and the group of 6. Dyads were much smarter; they spent more time adjacent to the group of 6. (The results were statistically significant.) Dyads of fish that had spent time together in the same tank (“familiar”) yielded similar results to dyads of fish that were not previously acquainted (“unfamiliar”).

Further experiments by Bisazza’s team hinted that the smarter fish led the way. In other words, merit-based leadership was responsible for dyads making better decisions than single fish alone.

If only humans were so smart.

Source:  Angelo Bisazza, Brian Butterworth, Laura Piffer, Bahador Bahrami, Maria Elena Miletto Petrazzini & Christian Agrillo. “Collective enhancement of numerical acuity by meritocratic leadership in fish.” Scientific Reports 4, Article number: 4560. Published 02-April-2014. doi:10.1038/srep04560