number of slackers drops almost to nothing, the group benefits, and cooperative behavior evolves in a strongly positive direction. And as the number of slackers declines, the risk to the punishers for punishing also drops. The altruistic punishers have made the group stronger, better able to survive, and have done so at a diminishing risk to themselves.
Now assume another situation: Everyone in the group is a rabid punisher. Common sense tells us that this is a toxic situation and the group suffers. An ideal group, then, has a certain percentage of altruistic punishers in it. In such groups, cooperation evolves rapidly.
In other words, one of our most treasured of human qualities — cooperation — evolved only because of the existence of punishment.
Bowles cites many psychology experiments that show human beings are avid to punish wrongdoing, even at expense to themselves. In one well-known study, college students were divided into pairs, A and B. A is given a hundred dollars and told that he can share as much or as little of that with B as he wishes. If B accepts the division, both get to keep the money. If B does not accept, both lose the money.
Logically, B would accept any amount of money from A — after all, it’s free money. Not so. B will gladly accept half and will almost always accept forty dollars. But when A offers B say, twenty dollars, B almost always refuses. Why? Because B wants to punish A for an unfair division, even though that also deprives B of money.
The experiment was extended. Now A shares money with B, with C as a witness. C has the option of punishing A if she thinks the division is unfair, but doing so costs C money. In chimp society, C wouldn’t give a damn about A and B’s sharing problem. But in human society, C will avidly punish A when the division starts to look “unfair,” even at a cost to herself.
Other experiments showed that when people punish, the dorsal striatum, a reward part of the brain, lit up. Those subjects who sacrificed the most to punish got the biggest charge from it.
Bowles mathematical simulations showed that an optimal society has a significant percentage of punishers. “There’s quite a bit of evidence,” he said, “that people really enjoy admonishing, inflicting harm on and punishing other people who are breaking social norms. They love to punish.” This, he points out, is a good thing. “A lot of the people who serve voluntarily in the military, or in criminal justice, are driven by motives of concern for other people. These are good people. If you look at history, what did liberal Europe create? It created a specialized group of people, wearing uniforms and badges, to enforce social norms” in a fair, evenhanded way.
As Bowles delved deeper into the mathematics of cooperation and punishment — particularly when he added warfare to the equation — a darker picture emerged, something he calls “parochial altruism.”
If you run the same simulations, but now set groups against each other in warfare, with the weaker groups experiencing extinction, the mathematics run toward a situation like this: Within the group, slackers are dealt with harshly by punishers. Punishment is even more important, because a slacker in war can seriously endanger the group. Cooperation and altruism evolve even more strongly. “Groups with lots of altruists,” said Bowles, “win wars.” The losers die; they don’t pass on their genes.
Let me pause to emphasize this disturbing point: Warfare in human history was essential for the evolution of cooperation and altruism.
“To call this controversial,” said Bowles dryly, “is an understatement.”
In the warfare scenario an additional, sinister effect becomes evident: Altruism within the group does not and must not extend to outside groups. The members of the other group must be demonized — otherwise, why would you kill them? They are bad to the bone and they must be punished. The altruism applies only within the
Janwillem van de Wetering