Forgiveness, Justice, and Tit-for-Tat

One of this blog’s devoted readers posted a reply the other day, chastising me for saying in relation to incompetent political leaders that I’d “get even” by ousting them from office. As a Quaker, he opposed retribution on principle.

I didn’t agree with him, but I filed his comment in the back of my mind, intending to think about it later. Then on Sunday, Anglican church services everywhere were reading a Biblical passage about forgiveness. The reader and the homilist both reminded me that Christ had ordered us to forgive, not just seven times, but seven times seventy times. Infinite forgiveness. Wow.

I am not sure what forgiveness means exactly, but I think it means to treat the offender as if he had not done anything wrong. If that’s the case, then I can’t do it — or can’t always do it — and don’t even believe in doing it. In fact, it seems that the social scientist in me conflicts with the spiritual aspirant in me, for social scientists plainly agree that “negative sanctions” are necessary ways of encouraging cooperation in all of us. Negative sanctions are, to put it bluntly, punishments. At a minimum, that would seem to include campaigning against someone like George W. Bush at the next election as retribution, not only for stupendous inompetence but for misleading the public and following policies that are contrary to the public interest. (I can do so because I have both Canadian and American citizenship.)

During the intercessions at my church we all have a chance to voice particular prayers. I said aloud that I was praying for wisdom — for the ability to pursue both justice and forgiveness, for I don’t always know how to do them both at the same time.

Afterward I chatted with Bill Whitla, a retired English professor and Anglican priest who had delivered the homily. How can we do both justice and forgiveness? I asked again. He replied that, yes, there has to be accountability. We have to treat people differently, according to their performances, but we should not resort to vengeance.

Good point, I guess. But what is vengeance? Isn’t it retribution? Isn’t it the imposing of “negative sanctions” — reciprocity for bad deeds? I may not hate another person, but I do feel obliged to reward or punish him for his excellence or willful malfeasance. I consider it my social obligation — my duty to uphold high standards of conduct.

About five years ago, for example, a woman stole some of my professional work. I had completed about half of the task of editing a book consisting of about 15 papers when she asked to join with me in the project as co-editor. Since I was pressed for time, I reluctantly agreed to this, knowing from past experience that her competence was limited and I’d have to double-check whatever she did. After a year or so, she sent the manuscript to the publisher, claiming to be the sole editor. I learned of it and protested to the publisher, who wouldn’t touch the manuscript, since I hadn’t even seen it since turning it over to her. Another publisher, however, did publish it in her name. I tried to bring her before the American Sociological Association’s ethics committee, but since she wasn’t a member, they had no authority. I could have sued, but this would have cost money and time, when I had more important things to handle. Hence, I let her get away with it.

Was this forgiveness? Objectively, yes, for she received no retribution. But was it a good thing? No. I actually feel guilty for failing to pursue the matter assiduously. I stood to benefit or lose nothing personally from the outcome; I had already retired, and there was no financial recompense involved. However, I believe that, to protect the integrity of one’s profession, one should never allow such unethical behavior to pass unremarked. (If you think you know this sociologist and want more details about her misconduct, contact me. Her initials are B.W.)

This brings me back to social science. My main career has been devoted to peace studies, and when I taught the course called “Negotiation and Nonviolence,” I always lectured on Prisoners’ Dilemma. In this famous game, two suspected partners in crime are arrested and confronted separately by the prosecutor, who gives them both a chance to confess and implicate the other. If they both remain silent, there will be too little evidence to convict either of them for their main offense, so they will get off with a light jail sentence for a minor infraction — six months. If both confess and betray each other, they will both serve two years. However — and here’s the kicker — if one stays silent and the other betrays him, the one who stays silent will serve ten years, while the one who betrayed him will go free.

Clearly, the optimum response for them, as a pair, is to stay silent and take the six-month sentence. However, if this is to work, they must both trust each other, for the penalty is extremely high for staying silent while the other partner “sings.” This is a “mixed motive” game, in that each prisoner has two incompatible interests — one as a member of the group, and another as an individual.

Prisoner’s Dilemma has been studied countless times. I used to divide my classes into two teams that had to decide separately how to respond. We’d play a whole series of games and it was common for teams to change their decision repeatedly, sometimes betraying the other side, sometimes cooperating with them. We’d keep score for each round, totalling up the sentences that each side amassed over a series of, say, ten rounds.

Peace researchers studied this problem: how to induce mutual trust and cooperation over time. Clearly, everyone learns from past experience. Whenever one prisoner “defects,” the other one must be dubious about trusting him in future rounds of play, but whenever cooperation has been displayed, one can rationally expect it to continue and therefore will play cooperatively.

Robert Axelrod’s book, The Evolution of Cooperation, is a classic study of Prisoners’ Dilemma, as played in a tournament by computers that were programmed to follow different strategies of decision-making. The winning strategy was called “Tit-for-Tat,” which had been developed by my friend Anatol Rapoport of the University of Toronto. According to this strategy, a player begins by cooperating with the other player (i.e. by staying silent instead of confessing and betraying the other). Then in each subsequent round, you play in the same way that the other side has just played. If he cooperates, you cooperate. If he betrays you, you betray him. This is perfect reciprocity — tit-for-tat. Of all the various strategies that played against each other in the tournament, this unforgiving one turned out to be most successful. Over time both sides tended to learn from experience that they’d be punished for betrayal and rewarded for cooperation. Players tended to develop more cooperative patterns of interaction. And if both sides played according to the Tit-for-Tat strategy, they would both receive minimum sentences.

However, there was one obvious pitfall. If either side made a mistake and betrayed the other on a single round, they would both get into a recurring pattern of mutual recriminations from which the strategy provided no possible exit.

I myself had some moral qualms about this Tit-for-Tat strategy too. I much prefer living according to a general principle of forgiveness, though whenever another person’s wrongdoing cannot be justified, I believe in imposing negative sanctions. In general, I prefer sanctions that require restitution, rather than punishment for its own sake. Restorative justice simply requires everyone to undo as much harm as possible from the effects of the wrongdoing.

Still, I couldn’t argue with the tournament’s results. Tit-for-Tat induced more cooperation over time than any strategy I would have chosen. Whenever we forgive or overlook the other side’s previous betrayal, we let him continue getting away with, and benefiting from, such acts. Consistent followers of Christ would be defeated (unless some hot-shot theologian can reconcile his principle of forgiveness with the principle of reciprocity).

This research still holds up. Negative sanctions are sometimes necessary.

However, Tit-for-Tat only works in a two-person game situation. In real life, we rarely play against each other in a series of identical situations lasting ten or twenty rounds. We usually have to decide whether to trust others whom we don’t know personally. Yet even here the principle of reciprocity applies, though it’s adapted to suit a large social group who interact somewhat anonymously.

This new principle is called the “indirect reciprocity.” I may not be able to punish you personally, since we’ll never do business together again, but I can pass the word on to others, so they will know whether to trust you. This process is the creation of “reputation,” and it’s becoming more important all the time in putting pressure on people to behave well. Some instances: Whenever you seek a job, the potential employer will probably phone someone who has worked with you and ask their appraisal of you. When you buy an item on e-Bay, you may refer to the past behavior of the person with whom you are dealing, and if her reputation is poor, you may choose not to buy. Likewise, when you buy a book from Amazon.com, you can count the stars it has been awarded by previous readers. When you read the articles posted on Slashdot.com, you’ll find that this is a moderated list; the articles are chosen by people who have previously established good reputations by posting articles themselves that have been rated highly.

Several writers — notably Hassan Masum —have been developing tools for calculating reputation efficiently. Masum writes, “Shared opinions drive society: what we read, how we vote, and where we shop are all heavily influenced by the choices of others. However, the cost in time and money to systematically share opinions remains high, while the actual performance history of opinion generators is often not tracked.” He and other like-minded researchers propose to make bring transparency to all our public dealings.

I’m a bit uncomfortable with this too, for I’m aware that reputations can take on a life of their own, whether or not they are accurate. I recall one instance involving a former mentor of mine who was a famous sociologist with a poor reputation that I considered unfair. In a small community, everyone will know everyone else and you can double-check the gossip you hear, but in a truly anonymous mass society, people simply pass on whatever impressions they have heard, with no chance to verify them.

Sometimes, of course, people contribute to the reputations of people they have never met. This seems a serious flaw in the process. If we’re going to take reputation into account, we need some way to establish the credibility of the persons on whose appraisals we depend. At a minimum, we want a system in which unfounded gossip doesn’t affect a person’s reputation. Fortunately, e-Bay and Amazon.com use methods in which the ratings are provided only by individuals who have personally experienced the conduct of the person being rated. This removes the chance that one’s reputation may be smeared by slander.

To make it work, the majority of users will need to participate in the rating process. I never used to do that when buying from Amazon or e-Bay, but now I see that I should take the trouble to do so. Reputation counts, and it should be fair and well-founded. But forgiveness? I don’t know how to bring that into the picture.