Kalyx Bowler

Proof in the Pudding

While cleaning out the old library at the Yerkes Primate Center, we unearthed forgotten treasures. One was the old wooden desk of Robert Yerkes, which is now my personal desk. The other was a film that probably had not been looked at for half a century. It took us a while to find the right projector, but it was worth the trouble. Lacking sound, the film had written titles inserted in between poor-quality black-and-white scenes. It featured two young chimpanzees working together on a task. In true slapstick style, befitting the movie’s flickering format, one of the chimps would slap the other on her back every time her dedication flagged. I have shown a digitized version to many audiences, causing much laughter in recognition of the humanlike encouragements. People are quick to grasp the movie’s essence: apes have a solid understanding of the advantages of cooperation.

The experiment was run in the 1930s by Meredith Crawford, a student of Yerkes. We see two juveniles, Bula and Bimba, pulling at ropes attached to a heavy box outside their cage. Food has been placed on the box, which is too heavy for one of them to pull in alone. The synchronized pulling by Bula and Bimba is remarkable. They do so in four or five bursts, so well coordinated that you’d almost think they were counting—“one, two, three … pull!”—but of course they are not. In a second phase, Bula has been fed so much that her motivation has evaporated, and her performance is lackluster. Bimba solicits her every now and then, poking her or pushing her hand toward the rope. Once they have successfully brought the box within reach, Bula barely collects any food, leaving it all to Bimba. Why did Bula work so hard with so little interest in the payoff? The likely answer is reciprocity. These two chimps know each other and probably live together, so that every favor they do for each other will likely be repaid. They are buddies, and buddies help each other out.

This pioneering study contains all the ingredients later expanded upon by more rigorous research. The cooperative pulling paradigm, as it is known, has been applied to monkeys, hyenas, parrots, rooks, elephants, and so on. The pulling is less successful if the partners are prevented from seeing each other, so success rests on true coordination. It is not as if the two individuals pull at random and, by luck, happen to pull together. Furthermore, primates prefer partners who cooperate eagerly and are tolerant enough to share the prize. They also understand that a partner’s labor requires repayment. Capuchin monkeys, for example, seem to appreciate each other’s effort in that they share more food with a partner who has helped them obtain it than with one whose help went unneeded. Given all this evidence, one wonders why the social sciences in recent years have settled on the curious idea that human cooperation represents a “huge anomaly” in the natural realm.

It has become commonplace to assert that only humans truly understand how cooperation works or know how to handle competition and freeloading. Animal cooperation is presented as mostly based on kinship, as if mammals were social insects. This idea was quickly disproven when fieldworkers analyzed DNA extracted from the feces of wild chimpanzees, which allowed them to determine genetic relatedness. They concluded that the vast majority of mutual aid in the forest occurs between unrelated apes. Captive studies have shown that even strangers—primates who didn’t know each other before they were put together—can be enticed to share food or exchange favors.

Despite these findings, the human uniqueness meme keeps stubbornly replicating. Are its proponents oblivious to the rampant, varied, and massive cooperation found in nature? I just attended a conference on Collective Behavior: From Cells to Societies, which addressed the extraordinary ways in which single cells, organisms, and entire species realize goals together. Our best theories about the evolution of cooperation stem from the study of animal behavior. Summarizing these ideas in his 1975 book Sociobiology, E. O. Wilson helped launch the evolutionary approach to human behavior.

Excitement about Wilson’s grand synthesis seems to have faded, though. Perhaps it was too sweeping and inclusive for disciplines that consider humans in isolation. Chimpanzees in particular are nowadays often depicted as so aggressive and competitive that they can’t be truly cooperative. If this applies to our closest relatives, so the thinking goes, we can justifiably ignore the rest of the animal kingdom. One prominent advocate of this position, the American psychologist Michael Tomasello, extensively compared children and apes, which has led him to conclude that our species is the only one capable of shared intentions in relation to common goals. He once condensed his view in the catchy statement “It is inconceivable that you would ever see two chimpanzees carrying a log together.”

This is quite an assertion, given Emil Menzel’s photographed and filmed sequences of juvenile apes recruiting one another to collectively prop a heavy pole up against the wall of their enclosure in order to get out. I have regularly seen chimps use long sticks as ladders to get across hot wire surrounding live beech trees; one chimp holds the stick while another scales it to reach fresh leaves without getting shocked. We have also videotaped two adolescent females who regularly tried to reach the window of my office, which overlooks the chimp compound at the Yerkes Field Station. Both females would exchange hand gestures while moving a heavy plastic drum right underneath my window. One ape would jump onto the drum, after which the other would climb on top of her and stand on her shoulders. The two females would then synchronously bob up and down like a giant spring; the one standing on top would reach for my window every time she came close. Well synchronized and clearly of the same mind, these females played this game often in alternating roles. Since they never succeeded, their common goal was largely imaginary.

Literally carrying a log together may not be part of these efforts, but this behavior is trained for all the time in Asian elephants. Until recently, the forest industry in Southeast Asia employed elephants as beasts of burden; now they are rarely used for this purpose anymore, but they still demonstrate their skills for tourists. At the Elephant Conservation Center near Chiang Mai, in Thailand, two tall adolescent bulls will effortlessly pick up a long log with their tusks, each standing on one end, draping their trunks over the log to keep it from rolling off. Then they will walk in perfect unison several meters apart, with the log between them, while the two mahouts on their necks sit chatting and laughing and looking around. They are most certainly not directing every move.

Training is obviously part of this picture, but one cannot train any animal to be so coordinated. One can train dolphins to jump synchronously because they do so in the wild, and one can teach horses to run together at the same pace because wild horses do the same. Trainers build on natural abilities. Obviously, if one elephant were to walk slightly faster than the other while carrying the log, or hold it at the wrong height, the whole enterprise would quickly unravel. The task requires step-by-step harmonization of rhythm and movement by the bulls themselves. They have moved from an “I” identity (I perform this task) to a “we” identity (we do this together), which is the hallmark of collective action. They end their performance by lowering the log together, moving it from their tusks into their trunks and then slowly to the ground. They set the heaviest log down on a pile without a single sound, impeccably coordinated.

When Josh Plotnik tested elephants on the cooperative pulling paradigm, he found a solid understanding for the need to synchronize. Teamwork is even more typical of group hunters, such as humpback whales, which blow hundreds of bubbles around a school of fish; the column of bubbles traps the fish like a net. The whales act together to make the column tighter and tighter, until several of them surface through its center with mouths wide open to swallow the bounty. Orcas go even further, in an action so astonishingly well coordinated that few species, including humans, would be able to match it. When orcas along the Antarctic Peninsula spot a seal on an ice floe, they reposition the floe. It takes lots of hard work, but they push it out into open water. Then four or five whales line up side by side, acting like one giant whale. They rapidly swim in perfect unison toward the floe, creating a huge wave that washes off the unlucky seal. We don’t know how the killer whales agree on the lineup or how they synchronize their actions, but they must be communicating about it before making their move. It is not entirely clear why they do it, because even though the orcas afterward carry the seal around, they often end up releasing it. One seal was deposited back onto a different ice floe to live another day.

On land, lions, wolves, wild dogs, Harris’s hawks (teams of which control the pigeons at London’s Trafalgar Square), capuchin monkeys, and so on, exhibit plenty of tight teamwork, too. The Swiss primatologist Christopher Boesch has described how chimpanzees hunt colobus monkeys in Ivory Coast: some males act as drivers, while others take up distant positions high up in a tree as ambushers waiting for the monkey troop to escape in their direction through the canopy. Since these hunts take place in the dense jungle of Taï National Park, and both the chimps and the monkeys are dispersed, it is hard to pinpoint what is going on in three-dimensional space, but it appears to involve role division and the anticipation of prey movement. The prey is captured by one of the ambushers, who potentially could quietly slink away with the meat but does exactly the opposite. During the hunt the chimps are silent, but as soon as a monkey is captured, they erupt in a pandemonium of hooting and screaming that draws everyone in, leading to a large cluster of males, females, and young jostling for position. I once stood under a tree (in a different forest) while this happened, and the deafening noise above me left little doubt about how highly chimps prize their meat. Sharing appears to favor hunters over latecomers—even the alpha male may go empty-handed if he failed to participate. The chimpanzees seem to recognize contributions to success. The communal feast that ensues is the only way to sustain this sort of cooperation, because why would anyone invest in a joint enterprise if not for the prospect of a joint payoff?

These observations obviously contradict the view that chimpanzees, and other animals, lack joint action based on shared intentions. One can imagine the head butting between two scientists with such diametrically opposite views as Boesch and Tomasello, who have offices in the same building. Was their appointment as codirectors of the Max Planck Institute in Leipzig an experiment on how human collaboration fares in the face of disagreement? Given these divergent perspectives, let me return to the experiments that led Tomasello to his human uniqueness claim. After testing both children and apes on a cooperative pulling task, he concluded that only the children exhibit shared intentionality.

The question of comparability has come up before, however, and fortunately there are photographs of the respective setups. One shows two apes in separate cages, each with a little plastic table in front of him that he can pull closer with a rope. Oddly, the apes do not occupy a shared space, as in Crawford’s classical study. Their cages are not even adjacent: there is distance and two layers of mesh between them—a situation that hampers visibility and communication. Each ape focuses on its own end of the rope, seemingly unaware of what the other is up to. The photo of the children, in contrast, shows them sitting on the carpeting of a large room with no barriers between them. They, too, are using a pulling apparatus, but they sit side by side in full view of each other and are free to move around, touch each other, and talk. These different arrangements go a long way toward explaining why the children showed shared purpose, and the apes did not.

Had this comparison concerned two different species—rats and mice, say—we would never have accepted such dissimilar setups. If rats had been tested on a joint task while sitting side by side and mice while being kept apart, no sensible scientist would permit the conclusion that rats are smarter or more cooperative than mice. We’d demand the same procedure. Comparisons between children and apes get exceptional leeway, however, which is why studies keep perpetuating cognitive differences that, in my mind, are impossible to separate from methodological ones.

In view of the ongoing controversy, we decided to move away from pair-wise testing—whether separate or together—and develop a more naturalistic setup. I sometimes refer to it as our proof-in-the-pudding experiment, since we sought to determine once and for all how well chimps handle conflicting interests: what happens to cooperation in the face of competition? The only way to see which tendency prevails is to provide an opportunity for the chimpanzees to express both at the same time.

My student Malini Suchak came up with the right apparatus to test a colony of fifteen chimps at the Yerkes Field Station. Mounted on the fence of their outdoor enclosure was an apparatus that required very precise coordination to be moved closer to obtain rewards: either two or three individuals had to pull at exactly the same time at separate bars. To coordinate with two partners was harder than with only one, but the apes had no trouble either way. They were sitting spaced out but in full view of one another. Since the whole group was present, there were many possible partner configurations. The apes could decide who to work with while also being on the alert for competitors, such as dominant males or females, as well as freeloaders who might steal rewards without doing any work. They could freely exchange information and freely choose partners, but also freely compete. No large-scale experiment of this kind had ever been tried.

If it is true that chimps can’t overcome competition, the test should produce total chaos! The colony should descend into a bickering bunch of apes, fighting over rewards and chasing one another away from the test site. Competitiveness should kill all shared objectives. I knew chimps long enough, however, that I didn’t worry much about the outcome of this test; I had studied conflict resolution among them for decades. Despite their poor reputation, I had seen too many scenes of chimpanzees trying to keep the peace and reduce tensions to worry that they would all of a sudden abandon such efforts.

Since Malini and the rest of us wished to see if the chimps could figure out the task on their own, she gave them no pretraining at all. All they knew was that there was a new apparatus and that food was associated with it. They proved remarkably quick learners, realizing that they had to work together and mastering both two-way and three-way pulls within days. Sitting next to one of the pull-bars, Rita would look up at her mother, Borie, who was asleep in a nest on top of a tall climbing frame. She’d climb up all the way to poke Borie in her ribs until she would come down with her. Rita would head for the apparatus, all the while looking over her shoulder to make sure Mom was following. Sometimes we had the impression that the chimps had reached an agreement without us knowing how. Two of them would walk side by side out of the night building, which is quite a distance away, and together head straight for the apparatus, as if they knew exactly what they were going to do. Talk about shared intentionality!

The main point of the study was to see if the apes would compete or cooperate. Clearly, cooperation won big time. We saw some aggression but virtually no injuries. Most fights were low level, such as pulling at someone to drag him or her away from the apparatus, chasing someone off, or throwing sand. Individuals also tried to gain access by grooming one of the pullers until this individual allowed them to take their spot. Cooperation at the apparatus went on almost nonstop, resulting in a total of 3,565 joint pulls. Freeloaders were avoided and occasionally punished for their activities, while overly competitive individuals quickly found out how unpopular their behavior made them. The experiment was conducted over many months, affording plenty of time for all the chimps to learn that tolerance paid off in terms of finding partners to work with. In the end, we found proof in the pudding that chimpanzees are highly cooperative. They have no trouble whatsoever regulating and dampening strife for the sake of achieving shared outcomes.

One possible reason that the behavior we observed was more in line with what is known from the natural habitat may be our colony’s background: by the time we tested them, our chimps had lived together for almost four decades. This is a long time by any standard, resulting in an unusually well-integrated group. But when we recently tested a newly formed group, in which many individuals had known one another for only a few years, we found the same high level of cooperation and low level of aggression. In other words, chimpanzees are generally good at conflict management for the sake of cooperation.

The current reputation of chimpanzees as violent and belligerent—“demonic” even—is almost entirely based on the way they treat members of neighboring groups in the wild: they occasionally carry out brutal attacks over territory. This fact has tainted their image, even though lethal combat is so rare that it took decades for scientists to agree on its occurrence. The rate of fatalities at any given field site is on average once every seven years. Moreover, it is not as if this behavior sets chimpanzees apart from ourselves. So why is it used as an argument against their cooperative nature, whereas in our own species intergroup warfare is rightly viewed as a collective enterprise? The same holds for chimpanzees—they almost never attack neighbors on their own. It is time for us to see them for what they are: talented team players who have no trouble suppressing conflicts within their group.

A recent experiment at the Lincoln Park Zoo in Chicago confirmed their cooperative skills. Scientists let a group of chimpanzees fish with dipsticks for ketchup that was stored in the holes of an artificial “termite” mound. At the beginning of the experiment, there were enough holes for all members to feed independently, but then the number of holes was reduced by one each day, until there were very few left. Since each hole was monopolizable, it was thought that the chimps would compete and squabble over access to the dwindling resource. But nothing of the kind happened. They adjusted to their new situation by doing the exact opposite: they peacefully gathered around the remaining holes—usually two at a time, sometimes in trios—dipping their sticks into them in alternation, each chimp politely awaiting his or her turn. Instead of a rise in conflict, all the scientists observed was sharing and turn taking.

Fishy Cooperation

Cooperation experiments often ask cognitive questions. Do the actors realize they need a partner? Do they know the partner’s role? Are they prepared to share the spoils? If one individual were to hog all the benefits, this obviously would imperil future cooperation. So we assume that animals watch not only what they get but also what they get compared to what their partner gets. Inequity is something to worry about.

This insight inspired an immensely popular experiment that Sarah Brosnan and I conducted with pairs of brown capuchins. After they performed a task, we rewarded both monkeys with cucumber slices and grapes after determining that they all favored the latter over the former. The monkeys had no trouble with the task if they received identical rewards, even if they both got cucumber. But they were vehemently opposed to unequal outcomes, if one got grapes and the other got cucumber. The cucumber monkey would contentedly munch on her first slice, but after noticing that her companion was getting grapes, she would throw a tantrum. She’d ditch her measly veggies and shake the testing chamber with such agitation that it threatened to break apart.

Refusing perfectly fine food because someone else is better off resembles the way humans react in economic games. Economists call this response “irrational,” since getting something is by definition better than getting nothing. No monkey, they say, should ever refuse food that she’d normally eat, and no human should reject a small offer. One dollar is still better than no dollar. Sarah and I are unconvinced that this kind of reaction is irrational, though, since it seeks to equalize outcomes, which is the only way to keep cooperation flowing. Apes may even go further than monkeys in this respect. Sarah found that chimpanzees sometimes protest inequity that goes the other way. They object not only to getting less than the other but also to getting more. Grape receivers may reject their own advantage! This obviously brings us close to the human sense of fairness.

Without going into further details, something encouraging happened in these studies. They were soon extended to other species, including outside the primates. It is always a sign of a field’s maturity when it expands. Researchers who applied inequity tests to dogs and corvids found reactions similar to those of the monkeys. Apparently, no species can escape the logic of cooperation, whether it involves the selection of good partners or the balance between effort and payoff.

The generality of these principles is best illustrated by the work on fish by Redouan Bshary, a Swiss ethologist and ichthyologist. For years Bshary has been enchanting us with observations of the interplay and mutualism between small cleaner wrasses and their hosts, the large fish from which the cleaners nibble away ectoparasites. Each cleaner fish owns a “station” on a reef with a clientele, which come and spread their pectoral fins and adopt postures that offer the cleaner a chance to do its job. In perfect mutualism, the cleaner removes parasites from the client’s body surface, gills, and even the inside of its mouth. Sometimes the cleaner is so busy that clients have to wait in queue. Bshary’s research consists of observations on the reef but also experiments in the laboratory. His papers read much like a manual for good business practice. For example, cleaners treat roaming fish better than residents. If a roamer and a resident arrive at the same time, the cleaner will service the roamer first. Residents can be kept waiting since they have nowhere else to go. The whole process is one of supply and demand. Cleaners occasionally cheat by taking little bites of healthy skin out of their client. Clients don’t like this and jolt or swim away. The only clients that cleaners never cheat are predators, which possess a radical counterstrategy: to swallow them. The cleaners seem to have an excellent understanding of the costs and benefits of their actions.

In a set of studies in the Red Sea, Bshary observed coordinated hunting between the leopard coral trout—a beautiful reddish-brown grouper that can grow to three feet in length—and the giant moray eel. These two species make a perfect match. The moray eel can enter crevices in the coral reef, whereas the trout hunts in the open waters around it. Prey can escape from the trout by hiding in a crevice and from the eel by entering open water, but it cannot get away from the two of them together. In one of Bshary’s videos, we see a coral trout and a moray eel swimming side by side like friends on a stroll. They seek each other’s company, with the trout sometimes actively recruiting an eel through a curious head shake close to the eel’s head. The latter responds to the invitation by leaving its crevice and joining the trout. Given that the two species don’t share the prey with each other but swallow it whole, their behavior seems a form of cooperation in which each achieves a reward without sacrificing anything for the other. They are out for their own gain, which they attain more easily together than alone.

The observed role division comes naturally to two predators with different hunting styles. What is truly spectacular is that the entire pattern—two actors who seemingly know what they are going to do and how it will benefit them—is not one we usually associate with fish. We have lots of cognitively high-level explanations for our own behavior and find it hard to believe that the same might apply to animals with much smaller brains. But lest one think that the fish are showing a simplified form of cooperation, Bshary’s recent work challenges this notion. Coral trout were presented with a fake moray eel (a plastic model capable of performing a few actions, such as coming out of a tube) that was able to help them catch fish. The setup followed the same logic as the pulling tests in which chimpanzees recruit help when needed, but not if they can complete the task alone. The trout acted in every way similar to the apes and were equally adept at deciding on their need for a partner.

One way to look at this outcome is to say that chimpanzee cooperation may be simpler than we thought, but another is to say that fish may have a better understanding of how cooperation works than we have been willing to assume. Whether all this boils down to associative learning by the fish remains to be seen; if it does, then any kind of fish should be able to develop this behavior. That seems doubtful, and I agree with Bshary that a species’s cognition is tied to its evolutionary history and ecology. Combined with field observations of cooperative hunting between coral trout and moray eels, the experiment suggests a cognition that suits the hunting techniques of both species. Since the trout takes most of the initiatives and decisions, it may all depend on the specialized intelligence of only one species.

These exciting excursions into nonmammals fit the comparative approach that is the hallmark of evolutionary cognition. There is no single form of cognition, and there is no point in ranking cognitions from simple to complex. A species’s cognition is generally as good as what it needs for its survival.

— Frans de Waal, Are We Smart Enough to Know How Smart Animals Are?