Everyone likes a brain-teaser, so let’s try one: You and your friend get busted for illegally downloading "Grown Ups 2" on your PC. The authorities lock you up, equally offended by your disregard for the law as by your taste for Kevin James sequels. But they’re willing to cut a deal with you, on the condition that you throw your friend under the bus as follows:
If you rat out your friend and testify against him, you walk free and your friend goes to prison for a year, assuming he stays silent. If you both testify against each other, you both go to prison for six months. However, if you both remain silent, you’re both out of prison in one month. What do you do?
This scenario is the prisoner’s dilemma, a staple of a field called game theory which tries to figure out why cooperative behavior exists and how individuals make decisions based on reason and self-interest.
According to classic game theory, a rational person in the scenario above would turn in his friend because it’s a cautious approach that ensures he won’t wind up in the worst-case scenario (staying silent while the friend sings, resulting in a one-year prison sentence). Everyone’s favorite "Beautiful Mind," John Nash, figured this out and got a Nobel Prize for it.
“But if that’s the rational behavior, why do we see cooperation in the real world?” asked Chris Adami, a computational biologist at Michigan State University.
The catch, Adami pointed out, is that in Nash’s prisoner’s dilemma, the prisoners aren’t allowed to talk to each other.
But if you introduce the possibility of communication, you can understand why the prisoners may act differently, making a deal to both remain silent and allowing cooperative behavior to evolve.
Despite the plausibility for cooperation to evolve, physicists William Press and Freeman Dyson recently came up with a model to show how selfish behavior should still dominate in nature.
Adami and his Michigan State colleague Arend Hintze set out to test the model by conducting a series of mathematical simulation games from the comfort of their own computers. (You don’t need to go to the Galapagos to study evolution!)
In their study, released Thursday in Nature Communications, they created theoretical populations of organisms in which some were selfish and some were “suckers,” as Adami put it. If the two personality types were unable to tell each other apart, the selfish individuals would attack one another and the suckers and eventually go extinct, and the suckers would win. However, if the selfish ones could recognize the suckers but not vice versa, then the selfish ones would win, playing nice with one another while killing off the suckers.
If each population could recognize the other, then the ultimate outcome would depend on the population size. For example, if the suckers had a numbers advantage, they could cooperate to overcome the selfish ones, allowing the cooperative behavior to quickly spread across the population and drive the selfish types to extinction. It wouldn’t pay to be a jerk.
So how does this all relate to the real world?
It’s not as simple as labeling everyone as “selfish” or “sucker” and predicting what will happen, but such models give us a simple way to understand how behaviors like cooperation and selfishness do -- or do not -- evolve. And there are opportunities to make the models more complex to better reflect the complexity of human behavior, Adami said.
Extending beyond humans, many animals demonstrate cooperative behavior, but being able to model that behavior and understand how it came about depends on recognizing how the animals communicate with one another, Adami said. Maybe it is time to go to the Galapagos after all.
Return to Science Now.
Follow on Twitter: @BradBalukjian