pmc.ncbi.nlm.nih.gov

How chimpanzees cooperate in a competitive world

Significance

Competitive tendencies may make it hard for members of a group to cooperate with each other. Humans use many different “enforcement” strategies to keep competition in check and favor cooperation. To test whether one of our closest relatives uses similar strategies, we gave a group of chimpanzees a cooperative problem that required joint action by two or three individuals. The open-group set-up allowed the chimpanzees a choice between cooperation and competitive behavior like freeloading. The chimpanzees used a combination of partner choice and punishment of competitive individuals to reduce competition. In the end, cooperation won. Our results suggest that the roots of human cooperation are shared with other primates.

Keywords: Pan troglodytes, freeloading, enforcement, punishment, partner choice

Abstract

Our species is routinely depicted as unique in its ability to achieve cooperation, whereas our closest relative, the chimpanzee (Pan troglodytes), is often characterized as overly competitive. Human cooperation is assisted by the cost attached to competitive tendencies through enforcement mechanisms, such as punishment and partner choice. To examine if chimpanzees possess the same ability to mitigate competition, we set up a cooperative task in the presence of the entire group of 11 adults, which required two or three individuals to pull jointly to receive rewards. This open-group set-up provided ample opportunity for competition (e.g., freeloading, displacements) and aggression. Despite this unique set-up and initial competitiveness, cooperation prevailed in the end, being at least five times as common as competition. The chimpanzees performed 3,565 cooperative acts while using a variety of enforcement mechanisms to overcome competition and freeloading, as measured by (attempted) thefts of rewards. These mechanisms included direct protest by the target, third-party punishment in which dominant individuals intervened against freeloaders, and partner choice. There was a marked difference between freeloading and displacement; freeloading tended to elicit withdrawal and third-party interventions, whereas displacements were met with a higher rate of direct retaliation. Humans have shown similar responses in controlled experiments, suggesting shared mechanisms across the primates to mitigate competition for the sake of cooperation.

Competition undermines joint efforts, even among relatives (1–5); therefore, its control is fundamental to the evolution and maintenance of cooperative relationships (6). Humans are considered exceptional in their capacity to favor cooperation over competition (7), and several alternative hypotheses have been proposed to explain how humans evolved to become uniquely cooperative (7–10). Human cooperation has been called a “huge anomaly” in the animal kingdom (7), whereas our closest relative, the chimpanzee, is used as a foil marked by limited cooperation, lacking in shared intentionality (11), and cognitively more adapted to competition than cooperation (12). These claims are curious, given how our best theories about cooperation derive from animal behavior and ultimately launched the evolutionary study of human behavior (e.g., ref. 13). It seems more parsimonious to assume that all cooperation, both human and animal, rests on fundamentally similar principles, including the need to mitigate competition and communicate shared goals (9). Our focus here is on the former principle.

Numerous enforcement mechanisms have been proposed to reduce competition in the context of cooperation, with punishment being granted the most attention (9). The use of enforcement mechanisms in response to competition suggests that competition is in violation of cooperative norms (14). In the laboratory, humans respond more cooperatively when punishment is allowed (15), and even the mere threat of punishment maintains high levels of cooperation (16). In many experiments, an uninvolved third party will punish a freeloader, at a cost to themselves, a phenomenon known as costly (or altruistic) punishment (15–17). Claims of ubiquity and human uniqueness in the use of punishment may have been overstated (9), however, because evidence of these strategies in field and cross-cultural studies of humans has been mixed (18). Punishment in cooperative contexts appears to depend on the scale of the society (19) and the possibility of retribution (20), whereas alternative enforcement mechanisms, such as partner choice, may be equally important for maintaining cooperation (21–23). In theoretical models, partner choice based on simple reinforcement, choosing the partner who has benefited you in the past, can lead to a cooperative equilibrium faster than if partner choice is not allowed (24). Partner choice, however, is hardly a uniquely human phenomenon (25, 26). This and various other enforcement strategies have been observed in species ranging from bacteria to vertebrates (reviewed in refs. 27 and 28). For example, legume plants sanction (punish) bacteria that do not fix nitrogen by decreasing the supply of oxygen to those bacteria (29), and client fish chase the much smaller cleaner wrasses, as soon as they cheat by taking a bite of skin, threatening the ultimate punishment (ingestion) (30).

Humans show flexible, facultative enforcement strategies adjusted to the context, the intention or motivation of the cheater, the availability of alternative strategies, and the scope of competition (23, 31). For example, humans only avoid partners that freeload if those partners cheated intentionally (32). As one of humans’ closest relatives, chimpanzees are excellent candidates for examining the potential for facultative enforcement outside of our species. Chimpanzees have gained the reputation of being overly competitive and violent, even though much of this reputation derives from intergroup encounters that, as in our own species, often turn violent (33–35). Moreover, because intergroup aggressive raids are often conducted jointly (36, 37), they hardly count as an argument against the cooperative nature of chimpanzees. Nevertheless, in most laboratory experiments on chimpanzees, cooperation has been rather limited, often severely constrained by competition and intolerance (2, 3, 38). Indeed, in one study cooperation failed to emerge without the experimenter “engineering” tolerant partnerships (3). Rather than viewing this as a failure to overcome competition, however, we find that chimpanzees in these restricted settings may not have had adequate opportunities to use appropriate enforcement strategies. Although a couple of studies have investigated the use of punishment (39, 40), there exist many other enforcement options to effectively mitigate competition. The natural behavior of chimpanzees is marked by highly cooperative behavior, such as complex group hunting with role division (41–43), territorial defense (36, 37), and “political” alliances during male power struggles (37, 44–46). Furthermore, patterns of decision-making during conflicts of interest in wild chimpanzees suggest that the potential for future cooperation limits competitive encounters (47). Given these observations, we expect this species to be capable of overcoming competition in favor of cooperation.

Carefully controlled laboratory environments, usually limited to pairs of chimpanzees, may actually curtail this species’ natural cooperative tendencies. This is why the present study has taken a different approach. Instead of bringing two or three apes together in a prearranged setting, we mimicked natural conditions by providing them with an open choice to select cooperation partners, plenty of ways to compete, but also to apply enforcement mechanisms in a flexible, facultative way. The set-up loosely mimicked a two- or three-person Stag Hunt game (48). In the Stag Hunt game individuals can choose to act individually for a smaller reward (analogous to individually hunting a hare) or cooperate for a larger reward (a stag). Competition results from acting on individual interests and likely results in fewer rewards than the adoption of a cooperative position at the apparatus. It is important to note that in the Stag Hunt game either all stag (cooperation) or all hare (individual interests) are equilibria. Although past experiments suggest that the equilibrium for chimpanzees is competition, or the hare response that favors individual interests over cooperation (2, 3, 38), the present study asks if this outcome is a result of the nature of chimpanzees or to the restrictions of traditional experimental settings.

A group of 11 chimpanzees was given a chance to pull cooperatively at an apparatus to obtain rewards. In half of the test sessions joint action by two chimpanzees was required to succeed, and in the other half joint action by three chimpanzees was required, increasing the difficulty of the task and adding complexity to the choice of partners. The chimpanzees were not trained on the task, but rather had to learn about the role of their partners through experience (see ref. 49 for more information about the task and a discussion of the rapid learning process). All testing took place in the open group in which the apes were free to choose their own partners. This set-up allowed ample opportunities for competition (Fig. 1). Dominant individuals could claim spots at the apparatus (i.e., through displacements), and bystanders could steal the food procured by the “workers” (i.e., freeloading).

Fig. 1.

Fig. 1.

The experimental setting. The cooperative pulling apparatus requires two or three chimpanzees to work together, but here a female in the foreground pulls in a crowd of five. These situations allow ample opportunity for the competition and freeloading analyzed in this study. Photograph by F.B.M.d.W.

We analyzed 94 h of video filmed during the cooperative task. The experiment took place in 1-h test sessions, occurring two to three times per week. During these sessions the chimpanzees could pull jointly at the apparatus to receive a reward, which is referred to as a successful cooperative act. Immediately following success, the apparatus was rebaited to allow them to try again. They could continue for as many successful joint pulls as possible within the 1-h test session. Throughout the test session the chimpanzees were free to approach or leave the apparatus and engage in competitive acts, like freeloading, displacement, and agonism.

Across the 94 test sessions, there were 3,656 successful cooperative acts and thousands of opportunities for competition. A previous analysis of a different subset of the data focused on the apes’ understanding of the task contingencies and how they selected partners with which to work (49), but this analysis ignored how they avoided or reduced competition when cooperating. Here we examine the latter strategies used in situ. If it is true that chimpanzees lack the enforcement mechanisms to mitigate competition without experimenter interference, the whole experiment should have dissolved into one bickering colony of apes. If, on the other hand, chimpanzees are natural team workers, as suggested by naturalistic observations, they should successfully cooperate in this environment.

Next, we compared this well-established group of chimpanzees with a newly formed one of 15 individuals. Specifically, we replicated the first 28 1-h test sessions (phase 1; see ref. 49), which was the number of sessions required to establish cooperation in the original group. The new group differed from our original study group in that they were engaged in a high level of competition outside the direct experimental context, because they were still competing on a daily basis for rank positions in the group. Regular observations outside the experimental sessions revealed a three-times higher rate of agonism compared with the first group. Furthermore, there were 10 rank reversals during this time in the new group, compared with zero in the original group. Thus, the global amount of competition in the second group was much higher, but the opportunities for competition (displacements and freeloading) in the experimental context remained the same. Given the importance of tolerance in establishing cooperation, a competitive group climate might negatively impact cooperation (50, 51). On the other hand, if chimpanzees are highly motivated to overcome competition and favor cooperation, we would expect the same strategies to mitigate competition related to the task in the newly formed group so that cooperation emerges regardless of group-level tensions.

Results

Overall Cooperation.

Despite more than 600 competitive interactions (freeloading, displacement, or agonism), the overall rate of cooperation was quite high, with the chimpanzees achieving 3,565 cooperative acts in 94 h (an average of one cooperative act every 1.54 min). A cooperative act was defined as a joint pull between two or three individuals at the apparatus that resulted in reward delivery. We created a “cooperation index” for each session by subtracting the number of competitive interactions from the number of cooperative interactions divided by the sum of both frequencies. This index can range from −1 (highly competitive) to +1 (highly cooperative). Across the 94 test sessions, the index significantly increased (Spearman’s ρ = 0.68, P = 0.001), showing that by the end of the experiment the chimpanzees had converged upon cooperation as their primary mode of interaction As Fig. 2 illustrates, the colony started out rather cooperatively, then became more and more competitive as the number of participants in the task increased for about 25 sessions. The chimpanzees soon left this competitive phase behind, however, growing ever more cooperative, reaching close to a +1 index by the end of the study. Overall, only 16.0% of sessions showed a negative cooperation index. At the individual level, only one individual had a negative cooperation index overall. This individual was Mai, an almost blind female of 48 y old, who never solved the cooperative task but freeloaded regularly, making for an index of −1. The other 10 group members, however, had an overall positive cooperation index ranging from 0.64 to 0.97.

Fig. 2.

Fig. 2.

The cooperation index, which runs from +1 (purely cooperative) to −1 (purely competitive), shows that over time the chimpanzees significantly began to favor cooperation over competition. Data displayed in 5-session blocks of 94 total test sessions.

Freeloading.

A freeloading event occurred when an individual tried to take food from a spot at which he or she did not pull at the apparatus. In the original group, we counted 175 freeloading attempts (2.1% of total rewards), 91 of which were successful (1.1% of total rewards). The success rate for freeloading (successes/attempts: 62.6%) was significantly higher than for cooperative pulling (44.0%) in a within-subject comparison (Wilcoxon signed-rank test: z = −2.13, n = 11, P = 0.03, two-tailed).

Each chimpanzee attempted to freeload at least twice (mean per individual, mean ± SD: 15.90 ± 16.03) and succeeded at least twice (mean ± SD: 8.27 ± 8.42). Both attempts and successful freeloading are considered “freeloading events” throughout. In 28.0% of freeloading events, there was no response from either the freeloader or the worker and the next trial proceeded as usual. In the remaining freeloading events, the chimpanzees used one or more enforcement mechanisms. A within-subjects Wilcoxon signed-rank test revealed that the chimpanzees were more likely to respond to a successful theft than a failed attempt (z = 2.20, P = 0.03) (Fig. 3). Specific enforcement strategies are discussed below and in Fig. 3. Costs of these strategies to the freeloader and the worker are described in Table 1.

Fig. 3.

Fig. 3.

Absolute frequencies of responses to successful and attempted thefts during all sessions. The chimpanzees were more likely to behaviorally respond to successful thefts, whereas they showed “no response” mostly when the freeloader was unsuccessful at actually obtaining the food.

Table 1.

Summary of counter strategies used by the workers and their costs to both the freeloader and the worker

Counter strategy by worker Cost to the freeloader Cost to the worker
Agonism Potential for injury Potential for injury
Protest Relationship damage? Potential for future freeloading, agonism
Withholding pulling Loss of future rewards Delay of rewards
Target withdraw None Loss of future reward opportunities
No response None Potential for future freeloading

Agonism and protest.

Workers protested by whimpering, pouting, screaming, or displaying a silent bared-teeth expression in 12.6% of freeloading events. Aggression, ranging from threats to hitting, biting, or grabbing, followed 9.7% of freeloading events, and was of relatively low intensity (noncontact; e.g., threats, swaying, bluffing). Freeloaders were significantly more likely to exhibit aggression than the targets of their freeloading (binomial, P = 0.02).

Third-party punishment.

Third-party outsiders intervened 14 times (8.0% of freeloading events), primarily responding to aggression between the freeloader and the worker (12 instances), with the remaining two times in response to protest. Four of these interventions were impartial, typical of policing among chimpanzees (44, 52). Nine of 10 partial interventions were in favor of the freeloading victim (binomial test: P = 0.005), with actions ranging from threatening or chasing off the freeloader to mild contact aggression (e.g., hitting). There were also four cases of consolation (where a third-party provided affiliative contact to a distressed party), all of which were in favor of the freeloading victim.

Partner choice.

A previously reported generalized linear mixed-model revealed that the chimpanzees relatively often approached partners who were kin or close in rank to themselves [dyadic sessions: Akaike Information Criterion (AIC) = 601.44, χ2 = 9.68, df = 0, P < 0.001; triadic sessions: AIC = 1199.22, χ2 = 4.12, df = 0, P < 0.001; 49]. Thus, the chimpanzees tended not to approach individuals who were much higher ranking than themselves unless they were relatives. Because chimpanzees with a greater dominance advantage over their intended victim were more likely to engage in freeloading (rank distance between freeloader and targeted worker correlated with freeloading rate: Spearman’s ρ = 0.25, n = 90 dyads, P = 0.004), the tendency to seek cooperation with similarly ranked partners may have helped deter freeloading. Additionally, two strategies emerged that allowed workers to avoid freeloaders following freeloading events. Workers withdrew following 24.6% of events, and stopped pulling until the freeloader moved away after 20.0% of events. Freeloaders withdrew following 13.7% of events, allowing the worker to continue working at the apparatus without threat of further interaction. An examination of time spent at the apparatus alone further emphasizes the role of avoidance. Mai, the one chimpanzee with a highly negative cooperation index, was alone 37.73% of the time. In contrast, the other individuals in the group spent between 12.57% and 29.55% of their time at the apparatus alone.

Displacements.

Displacements occurred more frequently and had a higher success rate (87.2% of attempts were successful) than freeloading (Wilcoxon signed-rank test by individual: z = 2.52, P = 0.01). Of displacements, 15.5% escalated into agonism (compared with 9.7% of freeloading) (Table 2). Chimpanzees tended to displace those lower in rank than themselves, and this tendency was greater the greater the rank distance between the two individuals (Spearman’s ρ = 0.31, n = 90 dyads, P < 0.001). Unlike responses to freeloading, the target of a displacement was more likely to attack the initiator (39 of 53 occurrences, binomial test: P = 0.002). Bystanders intervened in 6.3% of events, always partially but without a consistent preference. The initiator of a displacement was as likely to receive support as its target (10 vs. 11 occurrences). In contrast, consolation was significantly more likely to favor the target of the displacement than the initiator (binomial, P = 0.009).

Table 2.

Comparison of responses to behavior around the cooperation apparatus

Type of interaction Number of attempts Successful (%) Agonistic response (%) Who escalates response? Whom favored by intervening third-parties?
Cooperation 8,305 44.0 N/A N/A N/A
Displacement 335 87.2 15.5 Target Neither
Freeloading 175 52.0 9.7 Initiator Target

Cooperation Within a Newly Formed Group.

To test whether global levels of competition within a chimpanzee group affect the ability to establish and maintain cooperation, we replicated the first 28 sessions (phase 1) (49) with a newly established group. For comparison purposes, only the first 28 sessions in the original group are considered in this section. During the test sessions, the new group exhibited nearly three-times more agonism than the original group across all test sessions (Wilcoxon signed-rank comparison of the two groups by session: z = 2.91, n = 28, P = 0.004). However, the proportion of escalated (contact) aggression was similar for both groups (χ2 = 2.15, df = 1, P = 0.142). Despite frequent agonism overall, the new group was highly successful at cooperation, achieving 1,017 cooperative acts (approximately one cooperative act per every 1.62 min) compared with 765 by the original group in the same number of sessions.

This high level of participation in the cooperative task offered many opportunities for competition. There were significantly more displacements in the original group than in the new group (binomial, P = 0.02) but the rate of freeloading did not differ (binomial, P = 0.13). Responses were similar across both groups. In the new group, displacements escalated into agonism more frequently than freeloading (13.2% vs. 5.8% of events). The same suite of behavioral responses to freeloading was observed (protest: 11.8%; target withdraw: 17.6%; initiator withdraw: 17.6%) (Fig. 4). Third parties intervened on five occasions (14.7%). Three of these occasions provided consolation or support to the target, two to the initiator. The only case of agonistic support was directed against the initiator of freeloading. Both groups responded similarly to competitive acts; a 2 × 5 contingency test of the response categories (i.e., protest, agonism, target withdraw, initiator withdraw, and bystander involvement) showed no significant frequency differences between both groups (χ2 = 2.33, df = 4, P = 0.68).

Fig. 4.

Fig. 4.

Comparison of responses to freeloading between the two groups of chimpanzees: the original group and a newly established one.

Discussion

This study’s main outcome is that chimpanzees have absolutely no problem mitigating competition during a cooperative task: by the end of the experiment, cooperation prevailed. The chimpanzees cooperated almost nonstop, resulting in 3,656 cooperative acts in the first study. This result contrasts with earlier findings that chimpanzees are only able to cooperate in experimental environments orchestrated to minimize competition and maximize tolerance (2, 3). One possible explanation of the discrepancy with previous research is that we conducted many more sessions than in typical experiments. It is clear from the data that a cooperative equilibrium takes time to be achieved. Another possible explanation is, of course, the open-group setting of our experiment and options to enforce cooperation. Both differences bring our experiment closer to the conditions found in the natural habitat of chimpanzees.

Without interference from the experimenters, our chimpanzees exhibited approximately five cooperative acts for every one instance of competition, whereas a second group exhibited six cooperative acts for every competitive interaction. The “balance theory” of human relationships stipulates that stable relationships are typically characterized by a ratio of five positive acts for every one negative act (53, 54). For example, in a longitudinal study, couples who did not have at least a ratio of 5.0 for positive to negative interactions were more likely to get divorced (54). This theory stresses that it is not necessary to eliminate all conflict; rather, it is the balance between the two that is important. It is interesting that both groups of chimpanzees naturally achieved a similar balance, which led to a high degree of cooperative behavior.

In the context of the Stag Hunt game (see introductory paragraphs), it seems that the equilibrium that chimpanzees achieve may depend on the environment of the experiment (48). An analysis of cooperation in the original group revealed that there was initially an increase in competition as more and more individuals developed an understanding of the task (thus increasing opportunities for freeloading and displacements), followed by a sharp rise in cooperation during the second half of the experiment. Furthermore, the results suggest that it required more effort to cooperate rather than compete (as measured by attempts relative to successes). So, participating in a cooperative manner signaled a willingness to invest that effort to get rewards. This result also suggests that although the experiment was set up to test cooperation, competition easily arose within the experiment. Thus, although competition had the potential to thwart cooperation, over time the chimpanzees converged on cooperation as their primary strategy.

The chimpanzees used a variety of enforcement strategies, each of which had costs and benefits to the target (Fig. 3). These strategies included: (i) direct deterrence (i.e., agonism or protest), (ii) third-party punishment, and (iii) partner choice, which was the primary response of the targets. Our earlier research demonstrated that chimpanzees are more likely to approach individuals close in rank to themselves at the apparatus (49). Our current results suggest that rank advantage played a role in both displacements and freeloading, indicating that the chimpanzees may have been pre-emptively avoiding competition by avoiding distantly ranked individuals altogether. Once freeloading occurred, the chimpanzees had two choices to avoid future freeloading: withdrawal or withholding of pulling. Withdrawal may have specifically been playing a role in avoiding Mai, a chimpanzee who exclusively freeloaded and did not cooperate. She was left at the apparatus alone more frequently than everyone else in the group. Withholding of pulling allowed chimpanzees to maintain possession of their position at the apparatus (thus ensuring future rewards) without falling victim to further freeloading. Withholding of pulling occurred at a much lower rate, probably because of its short-term opportunity cost. If a chimpanzee does not proactively avoid a competitive individual by choosing not to work with them (partner choice), both withdrawal and withholding of pulling are alternatives to avoid allowing other individuals reap benefits for free (9).

Although relatively rare, third-party interventions were mainly directed against freeloaders, but not displacers. Third-party punishment of freeloaders and uncooperative partners is sometimes considered uniquely human (15–17). However, we observed a modest number of aggressive interventions aimed specifically at freeloaders. For example, a freeloader grabbed the target and began hitting her while attempting to pin her to the ground. Four chimpanzees intervened: one female impartially tried to get in between the two contestants; the other three bystanders, including the alpha male, directed aggression against the cheater. This aggression included threats, bluffing, and hitting. The fight was finally broken up with the alpha male charging through and hitting the freeloader on the head. Although intervening in a fight is qualitatively different from humans paying to punish freeloaders in economic experiments and was a relatively rare response in the present experiment, the pattern of favoring the victim of the theft is suggestive of group-enforced social norms against freeloading (14). Indeed, the dissimilar responses to displacements and freeloading may reflect different social norms regulating these two types of competition. Displacement is a common expression of dominance, so it may be that chimpanzees tolerate displacement but not freeloading.

Finally, significant differences in the overall level of competition across the two groups did not compromise the ability of the chimpanzees to establish and maintain cooperation. Responses to competition during the task remained remarkably similar across both groups, despite one group being very well established and the other being newly formed. Overall, our findings contrast sharply with reports of high levels of competition, low levels of tolerance, and harassment-induced food transfers among chimpanzees (38, 55). Given that in previous studies chimpanzees did not have a free choice of partners, it may be that partner choice is an essential mechanism that is preferred over punishment to enhance cooperation (23, 56). This theory is consistent with evidence that other species prefer withdrawal over punishment as an enforcement mechanism (30), and both theoretical and empirical evidence favoring partner choice over punishment in humans (1, 9, 21–23, 56). The chimpanzees’ preference for cooperation during this study demonstrates their ability to inhibit competition to increase long-term payoffs. For chimpanzees, like humans, the flexible use of a variety of responses, including bystander interventions, exerts costs on competition to incentivize cooperation.

Materials and Methods

Subjects.

The videos analyzed in this study were of two groups of chimpanzees living in social groups at the Field Station of the Yerkes National Primate Research Center (YNPRC) of Emory University. Group one consisted of 1 male and 11 females and had been together for more than 20 y. Group two consisted of 3 males and 12 females and was formed from small subgroups introduced 3 mo before the start of testing. Testing occurred in the outdoor compound (SI Materials and Methods) with the entire group present. During testing, subjects had access to the indoor sleeping quarters. All procedures were approved by Emory University’s Institutional Animal Care and Use Committee, protocol #YER-2000180–53114GA, before the commencement of the study. The YNPRC is fully accredited by the Association for Assessment and Accreditation of Laboratory Animal Care.

Apparatus.

The apparatus required one chimpanzee to remove a barrier in order for another chimpanzee to simultaneously pull in a tray baited with food (see ref. 49). Following the establishment of dyadic cooperation, a second barrier was added requiring three individuals to work together (SI Materials and Methods). Food rewards were one small piece of fruit per success. At any time, bystander chimpanzees were free to attempt to take the rewards from participants.

Behavioral Coding and Interrater Reliability.

Each trial was videotaped from two angles (a side and front view) using high-definition digital video cameras. Which chimpanzees solved the task successfully and which received rewards was confirmed from video. When an individual tried to take food from a spot at which he or she did not pull (e.g., freeloading), we recorded the behavior of the freeloader (the thief), the worker (the victim), and any bystanders involved in the event. We also recorded displacements when one individual was present at the apparatus and another individual took possession of the bar. Additionally we recorded all occurrences of agonism. A full ethogram can be found in SI Materials and Methods. As per the field standard, a second rater blind to the intention of the study coded 10% (∼9 h) of the videos, encompassing 398 cooperative acts. Interrater reliability was excellent for all behavioral measures (Cohen’s κ for occurrence of cooperative acts: κ = 1.0; occurrence of competition: κ = 1.0; attempts versus successes: κ = 1.0; nonagonistic responses: κ = 1.0; aggression: κ = 0.80; bystander behavior: κ = 0.88; displacements: κ = 0.88). Raw data files are available upon request. Whereas video footage is not for distribution, it can be made accessible through arrangement with the authors.

Analyses.

Nonparametric statistics were used because the data were not normally distributed. Because there was interdependence between dyads for the dyadic analyses (e.g., rank difference, kinship, and affiliation), the data were subjected to 10,000 permutations to obtain an exact P value. All analyses were run using SPSS 20.0 (IBM) and R statistical software (2012).

SI Materials and Methods

Housing and Husbandry.

The group of chimpanzees was housed in a 711-m2 outdoor compound that contained a large climbing structure and several enrichment items (barrels, tires, and so forth) and was adjacent to five indoor interconnected runs containing sleeping platforms and swings. The chimpanzees were fed two daily meals of fruits, vegetables, and grains at ∼ 0830 hours and 1500 hours and had access to water ad libitum. All food used in this study was supplemental to the chimpanzees daily intake and at no time was food or water restricted.

Apparatus.

This task used a cooperative pulling apparatus. The barriers were blocking the tray and thus two or three individuals were required to coordinate their behavior to obtain food. The barriers were connected to a steel rod that extended 20 cm into the chimpanzee compound. Pulling on the rod caused the barrier to drop down via a spring/pulley mechanism. Once the barrier was pulled down a second individual used a similar rod (also extending 20 cm into the compound) to pull in the tray. Once the tray was pulled in all of the way (∼30 cm) food rewards dropped into a funnel, which delivered the rewards into the compound near the chimpanzees sitting at the apparatus.

Behavioral Coding.

The following ethogram was used in this study. Both failed attempts and successful stealing were considered freeloading events. Note that some, but not all of these responses are mutually exclusive. For example, a target cannot both withdraw and withhold pulling during a single freeloading event, but could protest and withhold pulling. Responses marked as “freeloading only” were not possible following a displacement.

Behavior recorded Description
Types of competition
 Attempted freeloading Initiator attempts to take food from a spot at which she or he did not pull. Includes: displacing the target before she or he obtained the food, putting hand or mouth at the end of the food delivery funnel, scrounging for dropped food, chasing the target away, intimidating (e.g., pilo erection, bluff charge) or physically attacking the target (e.g., hitting).
 Successful freeloading Any of the above behaviors that result in the initiator obtaining the food.
 Displacement Initiator takes possession of a pull bar that was currently in the possession of another chimpanzee. The initiator must fully put their hand on the bar and pull to count as a displacement.
Responses
 Withdraw The target or initiator leaves the apparatus before the next trial
 Protest One or more of the following behaviors: pouting, whimpering, silent teeth-baring, or throwing a noisy temper tantrum
 Withholding pulling Target remains at the apparatus but does not pull until the initiator moves away. Freeloading only.
 Aggression, noncontact Mild aggression including bluffing, swaying, threatening
 Aggression, contact Escalated aggression including hitting, grabbing, biting
 No response The chimpanzee maintained his/her position and made no overt responses before the baiting of the next trial. Freeloading only.

Acknowledgments

We thank Victoria Horner, Darby Proctor, Zanna Clay, Harold Gouzoules, Sarah Brosnan, Monica Capra, and Philippe Rochat for helpful discussions; Julia Watzek for statistical help; and the Veterinary and Animal Care staffs at the Yerkes National Primate Research Center (YNPRC) for maintaining the health of our research subjects. This study was supported by the Living Links Center, Emory’s PRISM Program (NSF GK12 DGE0536941); Emory’s Dean’s Teaching Fellowship program; Emory’s FIRST program, NIH/National Institute of General Medical Sciences IRACDA Grant K12GM00680 (to M.W.C.); the Expanding the Science and Practice of Gratitude Project run by the Greater Good Science Center in partnership with University of California, Berkeley with funding from the John Templeton Foundation; the Canisius Earning Excellence Program; and National Institutes of Health's Office of Research Infrastructure Programs base grant to the Yerkes National Primate Research Center, P51OD011132.

Footnotes

The authors declare no conflict of interest.

References

  • 1.West SA, Pen I, Griffin AS. Cooperation and competition between relatives. Science. 2002;296(5565):72–75. doi: 10.1126/science.1065507. [DOI] [PubMed] [Google Scholar]
  • 2.Chalmeau R. Do chimpanzees cooperate in a learning task? Primates. 1994;35(3):385–392. [Google Scholar]
  • 3.Melis AP, Hare B, Tomasello M. Engineering cooperation in chimpanzees: Tolerance constraints on cooperation. Anim Behav. 2006;72(2):275–286. [Google Scholar]
  • 4.Petit O, Desportes C, Thierry B. Differential probability of “coproduction” in two species of macaque (Macaca tokeana, M. mulatta) Ethology. 1992;90(2):107–120. [Google Scholar]
  • 5.de Waal FBM, Davis JM. Capuchin cognitive ecology: Cooperation based on projected returns. Neuropsychologia. 2003;41(2):221–228. doi: 10.1016/s0028-3932(02)00152-5. [DOI] [PubMed] [Google Scholar]
  • 6.Frank SA. Perspective: Repression of competition and the evolution of cooperation. Evolution. 2003;57(4):693–705. doi: 10.1111/j.0014-3820.2003.tb00283.x. [DOI] [PubMed] [Google Scholar]
  • 7.Fehr E, Fischbacher U. The nature of human altruism. Nature. 2003;425(6960):785–791. doi: 10.1038/nature02043. [DOI] [PubMed] [Google Scholar]
  • 8.Gintis H, Bowles S, Boyd R, Fehr E. Explaining altruistic behaviour in humans. Evol Hum Behav. 2003;24(3):153–172. [Google Scholar]
  • 9.West SA, Griffin AS, Gardner A. Evolutionary explanations for cooperation. Curr Biol. 2007;17(16):R661–R672. doi: 10.1016/j.cub.2007.06.004. [DOI] [PubMed] [Google Scholar]
  • 10.Burnham TC, Johnson DDP. The biological and evolutionary logic of human cooperation. Anal Kritik. 2005;27(2):113–135. [Google Scholar]
  • 11.Moll H, Tomasello M. Cooperation and human cognition: The Vygotskian intelligence hypothesis. Philos Trans R Soc Lond B Biol Sci. 2007;362(1480):639–648. doi: 10.1098/rstb.2006.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Hare B, Tomasello M. Chimpanzees are more skillful in competitive than in cooperative cognitive tasks. Anim Behav. 2004;68(3):571–581. [Google Scholar]
  • 13.Wilson EO. Sociobiology: The New Synthesis. Harvard Univ Press; Cambridge, MA: 1975. [Google Scholar]
  • 14.Harms W, Skyrms B. Evolution of moral norms. In: Ruse M, editor. Oxford Handbook of Philosophy of Biology. Oxford Univ Press; Oxford: 2008. pp. 434–450. [Google Scholar]
  • 15.Boyd R, Gintis H, Bowles S. Coordinated punishment of defectors sustains cooperation and can proliferate when rare. Science. 2010;328(5978):617–620. doi: 10.1126/science.1183665. [DOI] [PubMed] [Google Scholar]
  • 16.Fehr E, Gächter S. Altruistic punishment in humans. Nature. 2002;415(6868):137–140. doi: 10.1038/415137a. [DOI] [PubMed] [Google Scholar]
  • 17.Fehr E, Fischbacher U. Third party punishment and social norms. Evol Hum Behav. 2004;25(2):63–87. [Google Scholar]
  • 18.Gächter S, Herrmann B, Thöni C. Culture and cooperation. Philos Trans R Soc Lond B Biol Sci. 2010;365(1553):2651–2661. doi: 10.1098/rstb.2010.0135. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Marlowe FW, Berbesque JC. More ‘altruistic’ punishment in larger societies. Proc Biol Sci. 2008;275(1634):587–590. doi: 10.1098/rspb.2007.1517. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Denant-Boemont L, Masclet D, Noussair C. Punishment, counterpunishment and sanction enforcement in a social dilemma experiment. Econ Theory. 2007;33(1):145–167. [Google Scholar]
  • 21.Barclay P. Strategies for cooperation in biological markets, especially for humans. Evol Hum Behav. 2013;34(3):164–175. [Google Scholar]
  • 22.Barclay P, Willer R. Partner choice creates competitive altruism in humans. Proc Biol Sci. 2007;274(1610):749–753. doi: 10.1098/rspb.2006.0209. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Barclay P, Raihani N. Partner choice versus punishment in human prisoner’s dilemma. Evol Hum Behav. 2016;37(4):263–271. [Google Scholar]
  • 24.Skyrms B. Trust, risk, and the social contract. Synthese. 2008;169:21–25. [Google Scholar]
  • 25.Schino G, Aureli F. Reciprocal altruism in primates: Partner choice, cognition, emotions. Adv Stud Behav. 2009;39:45–69. [Google Scholar]
  • 26.Bshary R, Grutter AS. Experimental evidence that partner choice is a driving force in the payoff distribution among cooperators or mutualists: The cleaner fish case. Ecol Lett. 2002;5(1):130–136. [Google Scholar]
  • 27.West SA, Griffin AS, Gardner A. Social semantics: Altruism, cooperation, mutualism, strong reciprocity and group selection. J Evol Biol. 2007;20(2):415–432. doi: 10.1111/j.1420-9101.2006.01258.x. [DOI] [PubMed] [Google Scholar]
  • 28.Clutton-Brock TH, Parker GA. Punishment in animal societies. Nature. 1995;373(6511):209–216. doi: 10.1038/373209a0. [DOI] [PubMed] [Google Scholar]
  • 29.Kiers ET, Rousseau RA, West SA, Denison RF. Host sanctions and the legume-rhizobium mutualism. Nature. 2003;425(6953):78–81. doi: 10.1038/nature01931. [DOI] [PubMed] [Google Scholar]
  • 30.Bshary R, Grutter AS. Asymetric cheating opportunities and partner control in a cleaner fish mutualism. Anim Behav. 2002;63(3):547–555. [Google Scholar]
  • 31.West SA, et al. Cooperation and the scale of competition in humans. Curr Biol. 2006;16(11):1103–1106. doi: 10.1016/j.cub.2006.03.069. [DOI] [PubMed] [Google Scholar]
  • 32.Liddell TM, Kruschke JK. Ostracism and fines in a public goods game with accidental contributions: The importance of punishment type. Judgm Decis Mak. 2014;9(6):523–547. [Google Scholar]
  • 33.Manson JH, Wrangham RW. Intergroup aggression in chimpanzees and humans. Curr Anthropol. 1991;32(4):369–390. [Google Scholar]
  • 34.Watts DP, Muller M, Amsler SJ, Mbabazi G, Mitani JC. Lethal intergroup aggression by chimpanzees in Kibale National Park, Uganda. Am J Primatol. 2006;68(2):161–180. doi: 10.1002/ajp.20214. [DOI] [PubMed] [Google Scholar]
  • 35.Wilson ML, Wrangham RW. Intergroup relationship in chimpanzees. Annu Rev Anthropol. 2003;32:363–392. [Google Scholar]
  • 36.Mitani JC. Cooperation and competition in chimpanzees: Current understanding and future challenges. Evol Anthropol. 2009;18(5):215–227. [Google Scholar]
  • 37.Muller MN, Mitani JC. Conflict and cooperation in wild chimpanzees. Adv Stud Behav. 2005;35:275–331. [Google Scholar]
  • 38.Hare B, Melis AP, Woods V, Hastings S, Wrangham R. Tolerance allows bonobos to outperform chimpanzees on a cooperative task. Curr Biol. 2007;17(7):619–623. doi: 10.1016/j.cub.2007.02.040. [DOI] [PubMed] [Google Scholar]
  • 39.Jensen K, Call J, Tomasello M. Chimpanzees are vengeful but not spiteful. Proc Natl Acad Sci USA. 2007;104(32):13046–13050. doi: 10.1073/pnas.0705555104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Riedl K, Jensen K, Call J, Tomasello M. No third-party punishment in chimpanzees. Proc Natl Acad Sci USA. 2012;109(37):14824–14829. doi: 10.1073/pnas.1203179109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Boesch C. Cooperative hunting in wild chimpanzees. Anim Behav. 1994;48(3):653–667. [Google Scholar]
  • 42.Boesch C. Cooperative hunting roles among taï chimpanzees. Hum Nat. 2002;13(1):27–46. doi: 10.1007/s12110-002-1013-6. [DOI] [PubMed] [Google Scholar]
  • 43.Mitani JC, Watts DP. Why do chimpanzees hunt and share meat? Anim Behav. 2001;61(5):915–924. [Google Scholar]
  • 44.de Waal FBM. Chimpanzee Politics: Power and Sex Among Apes, 25th Anniversary Edition. Harper & Row; New York: 2007. [Google Scholar]
  • 45.Nishida T. Alpha status and agonistic alliance in wild chimpanzees (Pan troglodytes schweinfurthii) Primates. 1983;24(3):318–336. [Google Scholar]
  • 46.Nishida T, Hosaka K. Coalition strategies among adult male chimpanzees of the Mahale Mountains, Tanzania. In: McGrew W, Marchant L, Nishida T, editors. Great Ape Societies. Cambridge Univ Press; Cambridge, UK: 1996. pp. 114–134. [Google Scholar]
  • 47.Wittig RW, Boesch C. Decision-making in conflicts of wild chimpanzees (Pan troglodytes): An extension of the relational model. Behav Ecol Sociobiol. 2003;54(5):1–32. [Google Scholar]
  • 48.Skyrms B. The stag hunt. Proc Natl Acad Sci USA. 2001;75(2):31–41. [Google Scholar]
  • 49.Suchak M, Eppley TM, Campbell MW, de Waal FBM. Ape duos and trios: Spontaneous cooperation with free partner choice in chimpanzees. PeerJ. 2014;2:e417. doi: 10.7717/peerj.417. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Cronin KA, Sánchez A. Social dynamics and cooperation: The case of nonhuman primates and its implications for human behavior. Adv Complex Syst. 2012 doi: 10.1142/S021952591250066X. [DOI] [Google Scholar]
  • 51.Cronin KA, van Leeuwn EJC, Vreeman V, Haun DBM. Population level variability in the social climates of four chimpanzee societies. Evol Hum Behav. 2014;35(5):389–396. [Google Scholar]
  • 52.von Rohr CR, et al. Impartial third-party interventions in captive chimpanzees: A reflection of community concern. PLoS One. 2012;7(3):e32494. doi: 10.1371/journal.pone.0032494. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Gottman JM. What Predicts Divorce? The Relationship Between Marital Processes and Marital Outcomes. Psychology Press; New York: 1994. [Google Scholar]
  • 54.Gottman JM, Levenson RW. Marital processes predictive of later dissolution: Behavior, physiology, and health. J Pers Soc Psychol. 1992;63(2):221–233. doi: 10.1037//0022-3514.63.2.221. [DOI] [PubMed] [Google Scholar]
  • 55.Stevens JR. The selfish nature of generosity: Harassment and food sharing in primates. Proc Biol Sci. 2004;271(1538):451–456. doi: 10.1098/rspb.2003.2625. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Barker JL, Barclay P, Reeve HK. Within-group competition reduces cooperation and payoffs in human groups. Behav Ecol. 2012;23(4):735–741. [Google Scholar]