An evolutionary approach is useful in making sense of the complex phenomena related to group cooperation. In terms of practical applications, recommendations based in an evolutionary perspective are congruent with those of traditional social psychology. A superordinate authority may be selected to help lead group members to make decisions that are consistent with the welfare of the community (Hardin, 1968). Another technique is to make people aware of their mutual interdependence. Allison and Messick (1985) found that when individuals were educated about the impact of their behavior on the availability of resources, they acted in more socially responsible ways.
Social psychological research indicates that humans form coalitions on the basis of virtually any commonality of interest, and change alliances quickly when interests diverge (Tafjel, 1982). Krebs and Denton (1997) provided evidence that cognitive mechanisms differentiating in-groups from out-groups evolved to categorize individuals. Information about in-group members is processed in systematically more favorable ways than information about out-group members (Linville, Fischer, & Salovey, 1989).
One might use the cooperative imperative based on the mutual self-interest of survival and interdependence. In an approach similar to the "We're all in it together" propaganda campaigns originating in the Great Depression, social psychologists could attempt to persuade others that all humans are in the "in-group." In his famous experiment with summer camp children, Sherif (1966) was able to create hostile factions from previously neutral children. Then, though the use of a superordinate goal, Sherif was able to reduce group hostilities, eventually leading to the integration of the groups as a friendly, cooperative, and unified whole.
Of course this will not be an easy struggle. In-group/out-group rivalries are readily apparent in tensions between ethnic and religious groups. Even in our technologically advanced society, distinctions between groups are life threatening in areas as "diverse" as such as the former Yugoslavia, Rwanda, Northern Ireland, Somalia, and the Middle East. It will be important to see if the increasing "globalization" of those in westernized industrial nations will lead to feelings of the mutual interest of the entire human species through mass-communication.
One recommendation would be to co-opt kin-recognition mechanisms, in order to extend potential altruism towards previous out-groups. This would require convincing parents to let their children socialize with those from other ethnicities and cultures from a very young age. Counteracting the automatic formation and retrieval of stereotypes for these groups could lead to a reduction in in-group/out-group based discriminations. Of course the exposure approach is already advocated by many of those seeking to reduce ethnic tensions. The kin-recognition argument could lend validity to their efforts, although it would require that exposure be pervasive and genuinely interactive rather than brief and superficial.
Arguments for an evolutionarily stable strategy (ESS) of cooperation to promote the "good of the species" are usually rejected because natural selection operates more effectively within breeding populations than between them. Group selection theories have been reanimated by David Sloan Wilson and Elliot Sober, who argue that selection can operate not only at the individual level, but at the group level (Horgan, 1996). Wilson (1997) holds that natural selection on individuals cannot explain the existence of altruistic behaviors. Altruists may be less fit than non-altruists within a single group, but groups of altruists are more fit than groups of non-altruists. While groups may not think as a unit, they may act as a unit and shift the competition from between-individuals to between-groups. This would seem to concur with the inherent in-group/out-group cognitive biases known in social psychology, such as the ultimate attribution error. Social psychologists concur that activation of group stereotypes, which may contain an evaluative component, precedes conscious control and happens whenever the discriminating features of a group are present (Tesser & Martin, 1996).
Morell (1996) echoes this argument with the honey bee example. He states that colonies operate as individual units or super-organisms, since only the queen reproduces, and workers' genes are passed along through her. Selection operates at the hive level, rather than that of the individual. Although this framework can be used to illustrate the value of cooperation within groups, its logic can ultimately be reduced to selection at the individual level. Because of the nature of reproduction in bees (diploid queen, haploid male drones and diploid but sterile female workers), workers maximize their own genetic fitness through cooperation. The offspring produced by the queen has an r of .75 with the other workers, which is higher than the r of .5 usually achieved in sexual reproduction. Although the specialization in tasks benefits the group, it resembles a caste system more than an egalitarian society. Even in social insect groups which have more than one queen in a hive, such as the epiponine wasp Brachygastra mellifica, workers are on average significantly more related to queens than to other workers (r of .37 versus .23) and should therefore suppress each other in order to let the queens produce the eggs. Though many workers have developed ovaries and could lay eggs, the genetic analyses showed that most or all males come from queens (Hastings, Queller, Eischen & Strassmann, 1998).
Although the existence of altruistic behaviors deriving from group-selection is a theoretical possibility, most scientists believe that few, if any species have met conditions necessary for group selection (Krebs, 1998). This would require between-group selection for altruism to occur at a faster rate than within-group selection for selfishness, and the chance of selfish individuals invading the group must be low. The newer group selection models mirror those of individual selection, so they can be mathematically translated into each other (Reeve, 1998). Assuming that the tendency to sacrifice oneself for the sake of one's group varies among individuals within groups, those with more selfish tendencies will survive better than their more altruistic neighbors. This would lead the group to eventually become more selfish in nature. In the case of harvester ants, victory in between-group competitions leads to fatal competition between queens of the victorious colonies (Reeve, 1998).
Cosmides (1989) has demonstrated cognitive mechanisms believed to have evolved to detect deception related to cheating in resource exchange. Since humans are the most cognitively advanced species, they will also have the most advanced forms of deception (Krebs, 1998). In addition to the previously mentioned cheating regarding mate selection criteria, individuals may pretend to give more than they take, when they actually take more than they give. Portraying oneself as an indiscriminate altruist can reduce discriminatory assistance in others, resulting in net gains (Alexander, 1987).
Of course, of the most effective methods of deception is when it coincides with self-deception. Self-deception would reduce or prevent the awareness of deception, hiding any subtleties that would provoke a cheater-detection alarm (Trivers, 1985). When dividing resources and calculating debts, individuals tend to believe that they deserve more than an objectively "fair" share, overvalue their contributions and undervalue the contributions of others (Fiske & Taylor, 1991).