Note: If you are seeing this page without navigation links on the left, Click Here.

Reciprocal Altruism

In addition to the prediction based on kin selection that differences in willingness of individuals to help others will be a function of their relatedness, the theory of reciprocal altruism predicts that altruistic behaviors will also be a function of beliefs about the recipient's likelihood of reciprocating (Trivers, 1971). The exchange of resources and support in times of need is adaptive since it benefits the viability of the group as a whole. As long as assistance is multidirectional, "altruistic" actions will occur since the donor may one day find themselves in need. The higher degree of social isolation prior to "modern times" would also increase the chances that other altruists will benefit from others' altruistic behavior, and the propagation of traits which encourage helping (Schroeder et al., 1995).

Reciprocal altruism and kin selection are the only mechanisms for the evolutionary maintenance of altruistic or beneficent traits whose theoretical bases have received widespread acceptance (Rothstein & Pierotti, 1988). Trivers (1971) stressed that reciprocal altruism is a phenomenon distinct from kin selection. Reciprocators are likely to establish long-term relations and to deliver most of their aid to other individuals genetically predisposed to reciprocation. Most acts of reciprocal altruism should involve indirect increments to inclusive fitness, at least in regards to alleles for reciprocation. Reciprocal altruism must involve aid that is returned to an original donor as a result of a behavior that has a net cost to an original recipient.

In prisoner's dilemma paradigm for non-related individuals, an individual would do best by defecting when the other person cooperates. With only one interaction, the evolutionarily stable strategy is to defect. If benefit for cooperation is greater than benefit for selfish defection, over multiple interactions cooperation will be evolutionarily stable. Also, cooperation will spread in a population of defectors if the consequences to the altruist are better than if both defected (Reeve, 1998).

Tit-for-tat (cooperate on first move, than mirror actions of the other) will be an evolutionarily stable strategy against defection if the probability for future interaction is sufficiently high. In game theory calculations, the probability of future interaction, w, corresponds to Hamilton's r. Those pursing the tit-for-tat strategies and unconditional altruists will be equally viable in a closed population. In case of a rare defector, tit-for-tat will win out unless there is a significant degree of genetic relationship, rb > c (Reeve, 1998).

In order for tit-for-tat to become established in a population of defectors, tit-for-tat strategists must meet each other with some minimum threshold frequency. Kin-indiscriminate tit-for-tat may be able to spread if it increases the number of partners and if kin-recognition errors are frequent enough that benefits are received from non-kin. Under harsh conditions, it is better to cooperate, while under mild conditions it is better to defect. Still, it may be easier to use tit-for-tat rather than trying to base actions on a judgment of the harshness of conditions. The spread of tit-for-tat may also be facilitated by social learning. An individual bases their cooperative tendency on the prior cooperative history of potential partners. Also, a superordinate enforcer may punish for mutual defection (Reeve, 1998).

In computer simulated prisoner's dilemma games, Axelrod and Hamilton (1981) found that if tit-for-tat reciprocity invades a system dominated by selfish individualism in sufficient strength, it can successfully overtake the selfish population. This is conditional on reciprocity being directed towards other reciprocators, and since it may take several interactions to differentiate the reciprocators from the individualists, a stable and sizable group of reciprocators is necessary. Trivers (1971) cited six conditions that favor lead to the evolution of reciprocity: long life spans, a high degree of mutual dependence, low rate of dispersal, ability to assist conspecifics in combat, and flexible dominance hierarchies. The ability to recognize individuals and remember previous encounters with those individuals would also be prerequisite. Trivers (1985) felt that humans would have met these conditions in their ancestral environment.

Trivers (1985) stated that humans evolved a sense of justice to guard against cheating in reciprocal relationships. Trivers distinguished two forms of cheating. Gross cheating is when the recipient does not reciprocate or reciprocates very little in comparison to the original donor. Subtle cheating occurs when both parties reciprocate, but one consistently less so than the other. Trivers believed that human altruism depends on perceptions of fairness and feelings of moralistic anger, guilt, gratitude and sympathy.

Anthropological studies with quantifications of behavioral observations have consistently shown that cooperative and conflictual interactions are best predicted from genealogical relatedness, even when ideology and lip service say otherwise (Betzig & Turke, 1986). For example, kin selection based on intermarriages predicted which Binumarien communities in New Guinea were more likely to be at war (Hawkes, 1983). Optimal discrimination theory describes the setting of the optimal acceptance threshold for kin and reciprocators when considering cooperation, in psychological terms minimizing type I and type II error. As the frequency of interaction with non-kin or non-reciprocators increases, the threshold should become more stringent. (Reeve, 1998).