In the PDG, participants are required to make a simple choice: to cooperate or compete with one another. If both participants cooperate, they do well (in our example, they each receive $5). If one competes while the other cooperates, the competitor thrives at the expense of the cooperator (in our example receiving $8 to the cooperator’s zilch). If both compete, they each get $2. From the perspective of maximizing self-interest, the rational thing to do is to compete. The rub, though, is that, as in arms races, the use of shared resources, intimate life, and business partnerships, the mutual pursuit of self-interest leads to worse joint outcomes.
A tit-for-tat strategy was submitted by Anatol Rapaport. It is disarmingly simple: It cooperates on the first round with every opponent. Then it reciprocates whatever the opponent did in the previous round. An opponent’s cooperation is rewarded with immediate cooperation. The tit-for-tat was not blindly cooperative, however: it met an opponent’s competition with competition. Defection was punished with immediate defection.
Axelrod held a second tournament that attracted the eager submission of sixty-two strategies. All of the entrants knew the results of the first round—namely, that tit-for-tat had won. All had the opportunity to return to their blackboard, to adjust their mathematical algorithms and carry out further computer simulations, and to devise a strategy that could unseat the tit-for-tat. In this second tournament, once again the tit-for-tat prevailed. The tit-for-tat did not prevail, it is important to note, against all strategies. For example, your more sinister mind might have anticipated that a strategy that starts out competitively and always competes will have the upper hand against the tit-for-tat, because it establishes an advantage in the first round (of course, this strategy scores few points, and suffers profoundly, against other purely competitive strategies). Overall, however, tit-for-tat, so simple and cooperative in its
Why tit-for-tat? Three principles underlie the tit-for-tat and also underlie emotions like compassion, embarrassment, love, and awe, which promote the meaningful life. A first is what might be called cost-benefit reversal. Giving to others is costly. Devoting resources to others—food, affection, mating opportunities, protection—entails costs to the self. In the long run, generosity risks dangerous exploitation if it is directed at others who do not reciprocate in kind. The costs of giving constrain the tendency toward cooperation.
Built into the human organism, therefore, must be a set of mechanisms that reverse the cost-benefit analysis of giving. These mechanisms might prioritize the gains of others over those of the self, and transform others’ gains into one’s own. The tit-for-tat instantiates this principle of cost-benefit reversal. Its default setting is to cooperate, to benefit the other as well as the self. It is not envious; the tit-for-tat does not shift strategy as its partner’s gains mount. And it forgives; it is willing to cooperate at the first cooperative action of its partner, even after long runs of mean-spirited defection.
The emotions that promote the meaningful life are organized according to an interest in the welfare of others. Compassion shifts the mind in ways that increase the likelihood of taking pleasure in the improved welfare of others. Awe shifts the very contents of our self-definition, away from the emphasis on personal desires and preferences and toward that which connects us to others. Neurochemicals (oxytocin) and regions of the nervous system related to these emotions promote trust and long-term devotion. We have been designed to care about things other than the gratification of desire and the maximizing of self-interest.
A second principle is what we might call the principle of reliable identification. This is clearly evident in the tit-for-tat—it is easy to read. There is no trickery to it, no Machivellian dissembling, no strategic misinformation. It would likely take only five to ten rounds against the tit-for-tat to make confident predictions about its future moves. Contrary to what you see on cable poker tournaments (where stone faces and inscrutability are the demeanor of the day), in the emergence of cooperative bonds transparency of benevolent intent is the wiser course. Cooperation is more likely to emerge and prosper when cooperative individuals can selectively interact with other good-natured individuals.