AI Quiz PDF

PROBLEMS OF GROUP LIVING 250 The winning strategy in “iterated prisoner’s dilemma” games is called tit for tat. Axelrod and Hamilton discovered this strategy by conducting a computer tournament. Economists, mathematicians, scientists, and computer wizards from around the world were asked to submit...

PROBLEMS OF GROUP LIVING 250 The winning strategy in “iterated prisoner’s dilemma” games is called tit for tat. Axelrod and Hamilton discovered this strategy by conducting a computer tournament. Economists, mathematicians, scientists, and computer wizards from around the world were asked to submit strategies for playing 200 rounds of the prisoner’s dilemma. Points were rewarded in accordance with the payof matrix shown in Figure 9.1. The winner was whoever had the highest number of points. The strategies consisted of decision rules for interacting with other players. Fourteen strategies were submitted and were paired in competition in a round-robin computer tournament. Some strategies were highly complex, involving contingent rules for modeling the other’s strategy and suddenly switching strategies midstream. The most complex had 77 lines of statements in the computer language Fortran. The winner of the tournament, however, employed the simplest strategy of all, tit for tat, containing a mere four lines of Fortran statements. It had two simple rules: (1) Cooperate on the frst move and (2) reciprocate on every move thereafter. In other words, start by cooperating, and continue cooperating if the other is also cooperating. If the other defects, however, then defect in kind. Trivers (1985) aptly labeled this “contingent reciprocity.” Axelrod (1984) identifed three features of this strategy that represented the keys to its success: (1) Never be the frst to defect—always start out by cooperating, and continue to cooperate as long as the other player does also; (2) retaliate only after the other has defected—defect immediately after the frst instance of non-reciprocation; and (3) be forgiving—if a previously defecting player starts to cooperate, then reciprocate the cooperation and get on a mutually benefcial cycle. To summarize: “First, do unto others as you wish them to do unto you, but then do unto them as they have just done to you” (Trivers, 1985, p. 392). Strategies for encouraging cooperation that will, in turn, lead to the success of tit for tat are discussed in Box 9.1. The results of this computer tournament suggest that cooperation can evolve fairly easily in nature. 9.1 Strategies for Promoting Cooperation According to Axelrod’s (1984) analysis of tit for tat as a key successful strategy, several practical consequences follow for the promotion of cooperation. First, enlarge the shadow of the future. If the other individual thinks that you will interact frequently in the extended future, he or she has a greater incentive to cooperate. If people know when the “last move” will occur and that the relationship will end soon, there is a greater incentive to defect and not cooperate. Enlarging the shadow of the future can be accomplished by making interactions more frequent and by making a commitment to the relationship, which occurs, for example, with wedding vows. Perhaps one reason that divorces ofen get nasty, marred by unkind acts of mutual defection, is that both parties know the “last move” and hence perceive a sharply limited shadow of the future. A second strategy that Axelrod recommends is to teach reciprocity. Promoting reciprocity not only helps oneself by making others more cooperative, it also makes it more difcult for exploitative strategies to thrive. The larger the number of those who follow a tit-for-tat strategy, the less successful one will be in attempting to exploit others by defecting. Essentially, the co-operators will thrive through their interactions with each other and the exploiters will sufer because of a vanishing population of those on whom to prey. A third strategy for the promotion of cooperation is to insist on no more than equity. Greed is the downfall of many, perhaps best exemplifed by the myth of King Midas, whose lust for gold backfred when everything he touched, even the food he wanted to eat, turned to gold. The beauty of tit for tat as a strategy is that it does not insist on getting more than it gives. By promoting equity, tit for tat elicits cooperation from others. A fourth strategy to promote cooperation is to respond quickly to provocation. If your partner defects on you, a good strategy is to retaliate immediately. This sends a strong signal that you will not tolerate being exploited, which might prompt future cooperation. A fnal strategy for promoting cooperation is to cultivate a personal reputation as a reciprocator. We live in a social world in which the beliefs others have about us—our reputations—determine whether they will befriend or avoid us. Reputations are established through one’s actions, and word about one’s actions spreads. Cultivating a reputation as a reciprocator will make others seek you out for mutual gain. A reputation as an exploiter will lead to social shunning. The combined efect of these strategies will create a runaway pattern of cooperation, in which those who were formerly exploiters are forced to rehabilitate their bad reputations by becoming co-operators themselves. In this way, cooperation will be promoted throughout the group. 9 COOPERATIVE ALLIANCES Cooperation Among Non-Humans Each species is somewhat unique in the adaptive problems it has confronted over the course of its evolution, but diferent species can arrive at similar solutions to common adaptive problems. It is instructive to examine non-human species to see whether they have evolved cooperation. We will start with the fascinating case of vampire bats and then look at chimps, who are phylogenetically closer to humans. Food Sharing in Vampire Bats Vampire bats got their name because their survival depends on the blood of other animals. They live in groups of up to a dozen adult females and their ofspring. The males leave the colony when they are capable of independence. Vampire bats hide during the day, but at night they emerge to suck the blood of cattle and horses. Their victims, of course, are not willing donors. Indeed, the horses and cattle ofen fick away the bats to prevent them from feeding. The bats’ ability to feed successfully increases with age and experience. One study found that 33 percent of the younger bats (under 2 years old) failed to get blood on any particular evening, whereas only 7 percent of the bats older than 2 years failed to feed (Wilkinson, 1984). How do the bats survive failed attempts to fnd food? Failure at feeding, in fact, can quickly lead to death. Bats can go without blood for only 3 days. As shown by the statistics above, however, failure is common; all bats fail at one point or another, so the risk of death due to starvation is a constant threat. Wilkinson (1984) discovered that the bats regularly regurgitate a portion of the blood they have sucked and give it to others in the bat colony, but not randomly. Instead, they give regurgitated blood to their friends, those from whom they have received blood in the past. Wilkinson showed that the closer the association between the bats—the more often they were sighted together—the more likely they were to give blood to each other. Only bats that were sighted in close proximity at least 60 percent of the time received blood from that compatriot. Not a single bat gave blood to another bat with whom he associated for a lesser period of time. In another part of the study, Wilkinson (1984) used a captive colony of vampire bats to explore additional aspects of reciprocal altruism. He experimentally deprived individual bats of food and varied the length of time of the deprivation. Wilkinson discovered that the “friends” tended to regurgitate blood more often when their friends were in dire need and close to starvation (e.g., 13 hours from death) than when they were in mild need (e.g., 2 days from death). He also found that the starved bats who received help from their friends were more likely to give blood to those who had helped them in their time of need. In sum, vampire bats show all the signs of having evolved reciprocal altruistic adaptations. Chimpanzee Politics Among the chimpanzees at a large zoo colony in Arnhem, the Netherlands, a chimp named Yeroen reigned as the dominant adult male (de Waal, 1982). He walked in an exaggeratedly heavy manner and looked larger than he really was. Only occasionally did he need to demonstrate his dominance, raising the hair covering his body on end and running full speed at the other apes, who scattered in all directions in response to his charge. Yeroen’s dominance extended to sexual activity. Although there were four adult males in the troop, Yeroen was responsible for nearly 75 percent of the matings when the females came into estrus. 251 PROBLEMS OF GROUP LIVING 252 As Yeroen grew older, however, things began to change. A younger male, Luit, experienced a sudden growth spurt and challenged Yeroen’s status. Luit gradually stopped displaying the submissive greeting to Yeroen, brazenly showing his fearlessness. Once, Luit approached Yeroen and smacked him hard. Another time, Luit used his potentially lethal canines to draw blood. Most of the time, however, the battles were more symbolic, with threats and blufs in the place of bloodshed. Initially, all the females sided with Yeroen, allowing him to maintain his status. Indeed, reciprocal alliances with females are essential to the maintenance of status—males defend the females against attack from other males and act as “peacemakers” in disputes; in return, the females support the males, aiding in the maintenance of their status. One by one, however, the females gradually began to defect and sided with Luit as Luit’s increasing dominance became apparent. After 2 months, the transition was complete. Yeroen had been dethroned and started to display the submissive greeting to Luit. The mating behavior followed suit. Whereas Luit achieved only 25 percent of the matings during Yeroen’s reign of power, his copulations jumped to more than 50 percent when he took over. Yeroen’s sexual access dropped to zero. Although ousted from power and lacking sexual access, Yeroen was not ready to retire. Gradually, he formed a close alliance with an upcoming male named Nikkie. Neither Yeroen nor Nikkie dared to challenge Luit alone, but together they made a formidable alliance. Over several weeks, the alliance grew bolder in challenging Luit. Eventually, a physical fght erupted. Although they all sustained injuries, the alliance between Nikkie and Yeroen triumphed. After this victory, Nikkie secured 50 percent of the matings. And because of his alliance with Nikkie, Yeroen now secured 25 percent of the matings, up from his previous dethronement level of zero. Although Yeroen never again attained dominant status, his alliance with Nikkie was critical in avoiding total banishment from mating. Reciprocally, Nikkie’s alliance with Yeroen was critical in attaining dominance over Luit. Cooperative alliances are central to the lives of chimpanzees. They engage in reciprocal exchange of grooming and food sharing, in which services given roughly equal the services received over the long run within dyadic relationships (Jaeggi, De Groot, Stevens, & Van Schaik, 2013). Moreover, males regularly solicit alliances with females, grooming them and playing with their infants. Without alliances with the females, males could never attain a position of dominance in the troop. As part of the bid for alpha status, a male will bite or chase a female if she is found associating with an opponent. When she is no longer associating with the opponent, the male will be extremely friendly toward her and her infants. This is a key strategy in the formation of chimpanzee alliances: Try to sever the alliances of one’s opponents and enlist their former allies. Through de Waal’s fascinating study of chimpanzee politics, we catch a glimpse of the complexities of the evolution of reciprocal altruism—alliances that form not just between males but also between the sexes. Cooperation and Altruism Among Humans Social Contract Theory The theory of reciprocal altruism predicts that organisms can beneft through cooperative exchange. There is one problem, however: Many potential exchanges do not occur simultaneously. “If I give you a beneft now, I must trust that you will reciprocate and give me a beneft at some later time. If you fail to reciprocate, then I have incurred a net cost.” In short, relationships involving reciprocal exchange are vulnerable to cheating—when people take a beneft without paying the cost of reciprocation (Cosmides & Tooby, 1992, 2005). In nature, opportunities for simultaneous exchange sometimes occur. “I can give you a piece of fruit that I gathered in exchange for a piece of meat that you hunted.” But in many contexts, there are opportunities for cooperation in which simultaneous exchange is simply not possible. “If you are being attacked by a wolf, for example, and I rush to your aid, you cannot at the same time repay me for the cost I incurred.” 9 COOPERATIVE ALLIANCES Evolutionary psychologists Leda Cosmides and John Tooby have developed social contract theory to explain the evolution of cooperative exchange in humans, with special attention to how humans have solved the problem of cheating. The possibility of cheating poses an ever-present threat to the evolution of cooperation. The reason is that cheaters have an evolutionary advantage over co-operators, at least under certain conditions. “If I take the benefts that you ofer but then fail to return the favor at a later time, I beneft twofold: I have gained benefts and avoided incurring the reciprocal costs.” For this reason, over evolutionary time, cheaters will thrive more than co-operators until the entire population consists of noncooperators. Reciprocal altruism can only evolve if organisms have a mechanism for detecting and avoiding cheaters. If co-operators can detect cheaters and interact only with like-minded co-operators, reciprocal altruism can gain a toehold and evolve over time. The cheaters will be at a disadvantage because they fail to beneft by entering into cooperative exchanges. What specifc problems do people have to solve to evolve mechanisms that motivate forming social contracts and avoiding the ever-present threat of cheaters? Cosmides and Tooby (1992) outlined fve cognitive capacities: Capacity 1: The ability to recognize many diferent individual humans. “If you give me a beneft and I get lost in a ‘sea of anonymous others’ (Axelrod & Hamilton, 1981), you will be vulnerable to being cheated. You must be able to identify me and remember me as distinct from all other people.” The ability to recognize many individuals might seem obvious, but this is only because humans are so good at it. One study showed that people can identify others whom they have not seen for up to 34 years, with a recognition rate of over 90 percent (Bahrick, Bahrick, & Wittlinger, 1975). This ability is located in a specifc area of the brain. People with a lesion in a specifc place in the right hemisphere develop a highly specifc defcit: an inability to recognize faces called “prosopagnosia” (Gardner, 1974). Humans are also especially good at recognizing other individuals solely by the way they walk (Cutting, Proftt, & Kozlowski, 1978). In sum, there is good scientifc evidence that humans have evolved a profcient ability to recognize many diferent individuals. Capacity 2: The ability to remember the histories of interactions with diferent individuals. This capacity breaks down into several diferent abilities. First, one must be able to remember whether the person with whom one has interacted was previously a cooperator or a cheater. Second, one must be able to keep track of who owes what to whom. This requires some sort of “accounting system” for keeping track of the costs you have incurred and the benefts you have received from a specifc individual. Failure to keep track of these histories of interaction will make a person vulnerable to being cheated. If you fail to keep track of how much you have given the other person, then you have no way of knowing whether the beneft later returned compensates adequately for the cost you have incurred. Capacity 3: The ability to communicate one’s values to others. If your friend fails to understand what you want, how can he or she provide the benefts you need? If you fail to communicate your anger to a defector, you might be vulnerable to future defections. Consider an example from de Waal’s (1982) study of chimpanzees. The study concerned Puist and Luit, who had a longstanding relationship of mutual helping when one was under attack. This happened once after Puist had supported Luit in chasing Nikkie. When Nikkie later displayed [aggressively] at Puist, she turned to Luit and held out her hand to him in search of support. Luit, however, did nothing to protect her against Nikkie’s attack. Immediately Puist turned on Luit, barking furiously, chasing him across the enclosure and even hitting him (de Waal, 1982, p. 207). Puist appears to be communicating to Luit her dissatisfaction with Luit’s failure to help in a time of need. Although such chimp communications are nonverbal, among humans, language can be used to supplement emotional expressions and other nonverbal behavior as the medium of communication of desires, entitlements, and distress about an unfulflled obligation. The phrases “you owe me,” “I need this,” “I am entitled to this,” and “I want this” represent ways in which humans communicate their values to others. 253 PROBLEMS OF GROUP LIVING 254 Capacity 4: The ability to model the values of others. The fip side of the coin to communicating your values is the ability to understand the values of others. If you can detect when people are needy and how they are needy, the beneft you provide can be tailored to that need. If I provide you with a piece of meat, failing to recognize that you are not hungry and have an ample supply of meat already, then the beneft I provide will not be worth much to you. By understanding the desires and needs of others, you can tailor your exchanges to maximize the beneft you provide, making the other person more indebted to you. Capacity 5: The ability to represent costs and benefts, independent of the particular items exchanged. Cosmides and Tooby (1989) argue that many animals exchange a delimited set of items, such as food and sex. Humans, however, can and do exchange an astonishing array of items—knives and other tools, meat, berries, nuts, fsh, shelters, protection, status, access to friends, assistance in fghts, sexual access, money, blow guns, information about enemies, help on term papers, photos, and video clips, to name but a few. For this reason, evolved mechanisms of social exchange cannot be prewired to represent (conceptualize) and negotiate for specifc items. We must be able to understand and cognitively represent the costs and benefts of a wide range of items. It is our general ability to represent costs and benefts of exchanges, not a specifc ability tied to particular items, that has evolved in humans. In sum, social contract theory proposes the evolution in humans of fve cognitive capacities to solve the problem of cheaters and engage in successful social exchange. Humans must be able to recognize other individuals; remember the history of interactions with them; communicate values, desires, and needs to others; recognize them in others; and represent the costs and benefts of a variety of items of exchange. Evidence for Cheater-Detection Adaptations To test social contract theory, Cosmides and Tooby conducted more than a dozen empirical studies on people’s responses to logical problems. Logic refers to the inferences one can make about the truth of one statement from the truth of other statements, independent of their content. If I assert “if P, then Q,” then once you fnd out that P is true, you logically infer that Q must also be true. Unfortunately, humans do not seem to be very good at solving logical problems. Imagine that in one room are a few archeologists, biologists, and chess players (Pinker, 1997, p. 334). None of the archeologists are biologists, but all of the biologists are chess players. What follows from this knowledge? More than 50 percent of college students surveyed conclude from this that none of the archeologists are chess players—clearly an invalid inference because the statement “all biologists are chess players” does not imply that no archeologists play chess. No participants in this study concluded that some chess players in the room are nonarcheologists, which is logically derivable from the premises. And roughly 20 percent claimed that no valid inferences can be drawn at all from the above premises, which is clearly wrong. Consider one type of logic problem (Wason, 1966). Imagine that four cards are lying on a table. Each card has a letter on one side and a number on the other, but you can see only one side. Now consider this: Which cards would you need to turn over to test the truth value of the following rule: “If a card has a vowel on one side, then it has an even number on the other side.” Turn over only those cars you would need to turn over to test the truth value of this rule. If you are like the majority of people in most studies, you would turn over the card with the “a,” or the “a” and the “2.” The “a” card is certainly correct. Because it is a vowel, if it had an odd number on the back it would mean the rule is false. The “2” card, however, yields no information relevant to testing the rule. Because the rule does not state that