Cooperation Depends on the Partner’s Behavior
Since cooperation necessarily involves two (or more) individuals, there is always a potentially weak link in the interaction. When making a decision to cooperate, individuals have to not only understand the task, but also determine whether their partner will be a good partner. Capuchins are sensitive to situations in which a partner might fail to cooperate. In the previous study, the rewards were always dispersed, such that each monkey could control “their” rewards. When rewards were clustered together in the middle of the tray, such that one individual could monopolize them, cooperation was much less common (de Waal & Davis, 2002). This was true from the first trial, indicating that their reticence to cooperate when rewards were clumped is not based on learning during the course of the experiment.
Capuchins are also sensitive to their partner’s actual behavior. If rewards are altered such that only one of the monkeys gets rewarded, the unlucky monkeys cooperated at higher rates when their partner shared the food bonanza. Moreover, the rates of food sharing were lower in a control in which the subject could pull the tray in by themselves (also for a single reward). This indicates that the subjects understood when their partner’s help was essential and rewarded their partners for their assistance (de Waal & Berger, 2000).
Finally, capuchins are sensitive to relative payoffs between themselves and others, and are more likely to cooperate with partners who do not dominate rewards. In this mutual barpull, the food rewards were dispersed, but varied between the two monkeys. In some conditions, the monkeys received the same high- or low-value food, while in others, one monkey received a high-value food while the other received a low-value one. Critically, in this test the monkeys were not separated from one another and the experimenter did not determine which monkey received which reward. Instead, the monkeys had to work out for themselves which one got the high-value reward and which one got the low-value reward (Brosnan, Freeman, & de Waal, 2006).
We found that cooperation increased when partners shared access to the higher-value food. When one monkey repeatedly claimed the higher-value food, their partner quit cooperating. Interestingly, this held true across all conditions, not just the inequitable ones. So if a monkey consistently took the high-value reward, their partner quit cooperating with them not only when the rewards were unequal, but also when the rewards were equal and high or equal and low. Thus, it seems that they were not reacting to the actual distribution of the rewards, but instead to their partner’s behavior.
It is also of note that in this study, the monkeys were willing to accept short-term inequity—getting the lower value of the two rewards—as long as the long-term outcomes were approximately equal. This implies that the monkeys were evaluating their interactions based on the long-term relationship rather than each individual interaction. This is important for cooperation in natural situations, as it is rare that individual outcomes are ever completely equitable. In fact, these monkeys normally react negatively to inequitable outcomes as compared to a partner, a topic discussed in depth in the following section.