by Morton Hunt
A particular focus of his research was the behavior of people in “mixed-motive situations,” such as labor-management disputes or disarmament negotiations, where each side seeks to benefit at the other’s cost yet has interests in common with, and does not want to destroy, the other. In the 1950s he studied such situations intensively in the laboratory by means of his own modification of the Prisoner’s Dilemma game.48 In Deutsch’s version, each player seeks to win imaginary sums by making one of two choices—with results that depend on which of two choices the other player makes at the same time. Specifically, Player 1 can choose either X or Y, and Player 2 simultaneously can choose either A or B. Neither, in deciding what to do, knows what the other is going to do, but both know that every combination of their choices— XA, XB, YA, and YB—has different consequences. Player 1, for instance, thinks: “If I do X and he does A, we each win $9—but if he does B, I lose $10 and he wins $10. What if I do Y? If I do, and he does A, I win $10 and he loses $10—but if he does B, we each lose $9.” Player 2 is confronted by similar dilemmas.
Since neither knows what the other is doing, each has to decide for himself what move might be best. But as in the original Prisoner’s Dilemma, logical reasoning doesn’t help; only if both players trust each other to do what is best for both will they choose X and A respectively, and each win $9. If either mistrusts the other or tries to do the best for himself without regard to the other’s welfare, he may win $10 while the other loses that much—but is equally likely to lose $10 while the other wins that much, or, along with the other player, lose $9.
Deutsch varied the conditions under which his student volunteers played so as to simulate and test the effects of a number of real-life circumstances. To induce cooperative motivation, he told some volunteers, “You should consider yourself to be partners. You’re interested in your partner’s welfare as well as your own.” To induce individualistic motivation, he told others, “Your only motivation should be to win as much as you can for yourself. You are to have no interest whatever in whether the other person wins or loses. This is not a competitive game.” Finally, to induce a competitive mind-set, he told still others, “Your motivation should be to win as much money as you can for yourself and also to do better than the other person. You want to make rather than lose money, but you also want to come out ahead of the other person.”
Usually, players made their choices simultaneously without knowing each other’s choice, but sometimes Deutsch had the first player choose and then transmit his choice to the second player, who would then make his choice. At other times, one or both players were allowed to change their choice when they heard what the other had chosen. And sometimes both were allowed to pass each other notes stating their intentions, such as, “I will cooperate, and I would like you to cooperate. That way we can both win.”49
As Deutsch had hypothesized, when the players were oriented to think of each other’s welfare, they behaved in a trusting fashion (they chose X and A)—and did the best, collectively, even though either one would have been the big loser if the other had double-crossed him. But when they were told to try to win the most and to best the other, each usually assumed that the other was also out to win at his expense and made choices that were good for only one and bad for the other, or bad for both.
An encouraging result, Deutsch has said, is that “mutual trust can occur even under circumstances in which the people involved are clearly unconcerned with each other’s welfare, provided that the characteristics of the situation are such that they lead one to expect one’s trust to be fulfilled.”50 That is the case when, for instance, one player is able to propose to the other a system of cooperation, with rules and penalties for infractions; or when one knows, before committing himself to a choice, what the other was going to do; or when one can influence the outcome for the other, with the result that it is not in the other’s interest to violate an agreement.
Deutsch’s use of the modified Prisoner’s Dilemma game was a seminal event in social psychology. It led to hundreds of similar studies by others who modified and varied the conditions of play in order to explore a range of other factors that encouraged either cooperative or competitive styles of conflict resolution.
Deutsch himself soon moved on to another game that he and a research assistant, Robert M. Krauss, constructed to investigate how threats affect conflict resolution. Many people, during conflicts, believe they can induce the other side to cooperate by making threats. Embattled spouses hint at separation or divorce in an effort to change each other’s behavior; management warns strikers that unless they come to terms it will close down the company; nations in conflict mass troops on the border or conduct weapons tests in the attempt to wrest concessions from the other side.
In Deutsch and Krauss’s Acme-Bolt Trucking Game there are two players, both “truck drivers,” one with the Acme Company, the other with the Bolt Company. This map represents the world in which they interact.
FIGURE 20
Which works better—toughing it out or cooperating?
Time is of the essence for each player. Quick trips mean profit; slow ones, loss. Each begins moving his truck at the same time and at the same speed (the positions appear on control panels), and each can choose to go by the circuitous route or the short one. The latter, although obviously preferable, involves a stretch of one-lane road that accommodates only one truck at a time. If both players choose that route at the same time, they reach a bumper-to-bumper deadlock and one or both have to back out, losing money. Obviously, the best course is for them to agree to take turns on the one-lane road, thus allowing both to make maximum and nearly equal profits.51
To simulate threat making, Deutsch and Krauss gave each player control of a gate at his end of the one-lane strip. Each player, when bargaining, could threaten to close his gate to the other’s truck unless the other agreed to his terms. The experiment consisted of twenty rounds of play in each of three conditions: bilateral threat (both players controlled gates), unilateral threat (only Acme controlled a gate), and no threat (neither player controlled a gate). Another important variable was communication. In the first experiment, the players communicated their intentions only by the moves they made; in a second one, they could talk to each other; in a third, they had to talk to each other at every trial. Since the goal of both players was to make as much money as possible, the total amount of money they made in twenty rounds of play was a direct measure of their success in resolving the conflict. The major findings:
—The players made the greatest profit (collectively) when neither could make a threat; fared less well in the unilateral threat condition; and, contrary to common belief, did worst when each could threaten the other. (Could our former belief in “mutual deterrence” as the way to avoid nuclear war have been an unthinkably expensive misjudgment from which, through luck, we did not suffer?)
—Freedom to communicate helped little toward reaching an agreement, particularly if each could threaten the other. Nor did the obligation to communicate if both could threaten, although it did if only one could.
—If the players were coached about communicating and told to try to offer fair proposals to each other, they reached agreement more swiftly than when not tutored.
—When both players could make threats, verbal communication following a deadlock led to a useful agreement more quickly than if they were allowed to communicate only before the deadlock. Apparently, becoming deadlocked was a motivating experience. —The higher the stakes, the more difficulty they had reaching agreement. —Finally, when the experiment was run by an attractive female research assistant instead of a male, the players—male undergraduates—acted in a more macho fashion, used their gates more frequently, and had significantly more trouble reaching cooperative agreements.52
The Acme-Bolt Trucking Game instantly became a classic, was widely cited, and won the prestigious AAAS award for social science research.* Like many another ground-breaking study, it was the target of criticism, much o
f which questioned whether the variables it was based on are found in real life. But with time that question has been fairly well settled. The notion that a conflict can be thought of as a problem, and approached by thinking “What is the best way for us to solve it?” has been borne out by many other studies and has been turned into a number of programs of practical training. In 1986 Deutsch founded the International Center for Cooperation and Conflict Resolution at Teachers College, and this institute, the Program on Negotiation at Harvard Law School, the Conflict Resolution Consortium at the University of Colorado, and other similar centers have had considerable success in teaching constructive methods of settling disputes to negotiators for management and labor, divorce and corporate lawyers, government officials and legislators, teachers and students, tenants and landlords, family members, and others in conflict situations. If unresolved conflict is all too rife in our world, it’s because all too few embattled individuals and peoples know about—or care about—peaceful resolutions of their disputes.
Research on the topic continues. Heidi Burgess, co-director with Guy Burgess of the Colorado Consortium, says that currently the areas of special interest are “the way people frame conflicts” and how this affects “the way the conflict process is conducted and/or resolved” (thus carrying on Deutsch’s original work), and, branching out to other aspects of the field, “the impact of humiliation, anger, fear, and other strong emotions on conflicts and their resolution, the social-psychological effects of trauma, and approaches to trauma healing.”53
Attribution
In the 1970s, cognitive dissonance was displaced as the leading topic of social psychology by a new subject, attribution. The term refers to the process by which we make inferences about the causes of events in our lives and the behavior of others.
Our attributions, whether correct or incorrect, are more responsible than objective reality for how we think, what we feel, and how we behave. Studies have shown, for example, that we commonly attribute greater warmth, sexiness, and other desirable traits to good-looking people than to homely people, and behave toward them accordingly. Again, those who ascribe women’s lower employment status and pay scales to their fear of success and lack of assertiveness treat them differently from those who believe the causes are male prejudice, male dominance in the workplace, and traditional attitudes about woman’s proper role. All these are examples of what social psychologists call the “fundamental attribution error”—namely, “the strong tendency to interpret other people’s behavior as due to internal (dispositional) causes rather than external (situational) ones.”54
The phenomenon of attribution is captured in an old joke. Two men, one a Protestant and the other a Catholic, see a priest entering a brothel. The Protestant smiles sourly at the evidence of the hypocrisy of Catholics, the Catholic smiles proudly at the evidence that a priest will go anywhere, even into a brothel, to save the soul of a dying Catholic.
For those who prefer a more serious example, attribution is illustrated by an early experiment conducted by two former students of Lewin’s, John Thibaut and Henry Riecken. They assigned naïve volunteers, one at a time, to work on a laboratory project, in the course of which each realized that he needed the help of two other people present, one a graduate student, the other a freshman. (Both were accomplices of the researchers.) Each volunteer sought their help and eventually got it. When the volunteers were later asked why they thought the others had helped them, most said the graduate student had helped because he wanted to, the freshman because he felt obliged to. These attributions were based not on anything they had experienced but on the volunteers’ preconceptions about social status and power.55
Much other research has examined an extremely serious form of attribution error—the reasons given by people as to why other people tolerated or committed acts of hatred against groups and even accepted genocide of the hated people. A 2003 study asked Jewish and German visitors to Anne Frank’s home in Amsterdam, now a museum, whether the behavior of Germans during the Holocaust was due to their aggressive nature (an internal cause) or to the historical context in which the events occurred (external factors). By a considerable margin, the Jewish respondents attributed the German behavior to German aggressiveness, the German respondents to external factors (thus more or less absolving themselves of inner evil).56
Fritz Heider, an Austrian psychologist, had suggested the concept of attribution as early as 1927, but little notice was taken of it for many years. In 1958, Heider, who had long since immigrated to the United States, broadened the concept, proposing in his Psychology of Interpersonal Relations that our perceptions of causality affect our social behavior, and that we respond not to actual stimuli but to what we think caused them. An example: If a wife is trying to annoy her husband by not talking to him, he may think either that she is worried or that he has done something to offend her, and his actions will depend not on the real reason for her behavior but on what he attributes it to.57 Heider also made a valuable distinction between those attributions which point toward external causes and those which point toward internal ones. This preceded by eight years Julian Rotter’s important work on the attribution of internal versus external locus of control as a key personality trait.
Psychologists found Heider’s ideas exciting, since knowledge of the factors that lead people to make attributions would greatly increase the predictability of human behavior. Interest in attribution grew throughout the 1960s, and by the 1970s it had become one of the hot topics in social psychology.
But more a topic than a theory; indeed, it was a mass of small theories, each a reworking in attributional terms of some previous explanation of a sociopsychological phenomenon.58 Cognitive dissonance was reinterpreted as the self-attributing of one’s behavior to what one supposed one’s beliefs and feelings must be. (If circumstances compel me to behave badly toward someone, I tell myself that he deserves it and attribute my behavior to my perception of his “real” nature.) The foot-in-the-door phenomenon was similarly explained anew: if I give a little to a fund raiser the first time, and therefore give more a second time, it is because I attribute the first donation to my being a good and kindly person. And so on. Large areas of the territory of social psychology were invaded and laid claim to by the attributionists.59
More important than the reinterpretation of previous findings was the multitude of new discoveries resulting from attribution research. A few notable examples:60
—Lee Ross and two colleagues asked pairs of student volunteers to play a “quiz show game.” One was appointed questioner, the other contestant. Questioners were asked to make up ten fairly difficult questions to which they knew the answers, then pose them to the contestants. (Contestants averaged about six correct answers.) Afterward, all participants were asked to rate one another’s “general knowledge.” Nearly all the contestants said they considered questioners more knowledgeable than themselves; so did impartial observers of the experiment. Even though they knew that questioners had asked questions they knew the answers to, they attributed superior general knowledge to them because of the role they had played.
—Investigators discovered that we commonly attribute the behavior of highly noticeable, different-looking, or strikingly dressed people to inherent qualities, and the behavior of forgettable or ordinary-looking people to external (situational) forces.
—People’s reactions to the poor, alcoholics, accident victims, rape victims, and other unfortunates were explained in terms of the “just world hypothesis”—the need to believe that the world is orderly and just, and that it rewards us according to our deserts. This leads to the attribution of victims’ misfortunes to their own carelessness, sloth, risk taking, seductiveness, and the like. Some studies have found that the worse the plight of the victim, the more he or she is seen as responsible for it.
—Male college students were asked by the psychologist Stuart Valins to look at slides of nude women and rate their attractiveness. While looking at them, each man, through
earphones, heard what was supposedly his own heartbeat but was in reality recorded sound controlled by Valins. The lub-dub, lub-dub the volunteers heard was speeded up when they looked at certain slides but not others. When they later rated the appeal of the women, they named as particularly attractive those who seemed to have caused their heartbeat to speed up.
—Volunteers given false reports of how well they had done on tests tended to attribute supposed success to their own efforts or abilities, supposed failure to external causes such as the unfairness of the test, distracting noises, and so on.
—Researchers asked a group of nursery school children who had previously enjoyed drawing with multicolored felt-tip pens to play with them in order to receive Good Player awards. They asked a control group to play with the pens but said nothing about an award. Some time later, both groups were given access to the pens during free-play periods. The children who had received awards were much less interested in using them than the no-award group. The attributional interpretation: children who had expected a reward implicitly thought, “If I do it for the reward, I must not find drawing with it very interesting.”