Are We Smart Enough to Know How Smart Animals Are

by Frans de Waal

  Tied to the immense predation pressure under which these animals live is their ability for camouflage. Perhaps their most astonishing specialization, it provides an inexhaustible “magic well” for those who study them. The octopus changes color so rapidly that it out-chameleons the chameleon. Roger Hanlon, a scientist at the Marine Biological Laboratory in Woods Hole, Massachusetts, has collected rare underwater footage of octopuses in action. All we see at first is a clump of algae on a rock, but hidden among it is a large octopus indistinguishable from its surroundings. When the approaching human diver scares the animal, it turns almost white, revealing that it represented almost half the clump of algae. It speeds away while shooting a dark cloud of ink, which is its secondary defense. The animal then lands on the sea floor and makes itself look huge by spreading all its arms and stretching the skin between them into a tent. This frightening expansion is its tertiary defense.

  When this video clip is slowed down and played backward, it is easy to see how superb the original camouflage was. Both structurally and color-wise, the large octopus had made itself look exactly like an algae-covered rock. It did so by making its chromatophores (millions of neurally controlled pigment sacs in its skin) match their surroundings. But instead of exactly mimicking its background, which is impossible, it did so just well enough to fool our visual system. And it probably did more than that, since the octopus also takes other visual systems into account. Humans see no polarized or ultraviolet light and don’t have great night vision, whereas the octopus’s camouflage needs to trick all these visual capacities. In doing so, it draws on a limited set of patterns that it has in stand-by mode. Turning on one of these “blueprint” patterns allows it to blend in in a fraction of a second. The result is an optical illusion, but one realistic enough to save its life hundreds of times.28

  Sometimes an octopus mimics an inanimate object, such as a rock or plant, while moving so slowly that one would swear it is not moving at all. It does so when it needs to cross an open space, an activity that exposes it to detection. Imitating a plant, the octopus waves some of its arms above itself, making them look like branches, while tiptoeing on three or four of its remaining arms. It takes tiny little steps in line with the water movements. If the ocean is wild, plants sway back and forth, which helps the octopus disguise its steps by swaying in the same rhythm. On a waveless day, on the other hand, nothing else moves, so the octopus needs to be extremely careful. It may take twenty minutes to cross a stretch of sea floor that it otherwise might have crossed in twenty seconds. The animal acts as if rooted to the spot, counting on the fact that no predator will take the time to notice that it is actually inching forward.29

  The champion of camouflage, finally, is the mimic octopus, a species found off the coast of Indonesia that impersonates other species. It acts like a flounder by adopting this fish’s body shape and color as well as its typical undulating swimming pattern close to the sea floor. The repertoire of this octopus includes adopting the likeness of a dozen local marine organisms, such as lionfish, sea snakes, and jellyfish.

  We don’t know exactly how octopuses achieve this astonishing range of mimicry. Some of it may be automated, but there is probably also learning involved based on observations of other creatures and adoption of their habits. As primates, we find it impossible to relate to these remarkable capacities, and we may hesitate to call them cognitive. We tend to view invertebrates as instinct machines, arriving at solutions through inborn behavior. But this position has become untenable. There are too many remarkable observations—including the deceptive tactics of cuttlefish, close relatives of the octopus.

  Male cuttlefish courting a female may trick rival males into thinking there is nothing to worry about. The courting male adopts the coloring of a female on the side of his body that faces his rival, so that the latter believes he is looking at a female. But the same male keeps his original coloring on the female’s side of his body in order to keep her interested. He thus courts her surreptitiously. This two-faced tactic, called dual-gender signaling, suggests tactical skills of an order that we might expect in primates but not mollusks.30 Hanlon rightly claims that cephalopod truth is stranger than fiction.

  Invertebrates will probably continue to challenge students of evolutionary cognition. Being anatomically quite different yet facing many of the same survival problems as the vertebrates, they offer fertile grounds for convergent cognitive evolution. Among the arthropods, for example, we find jumping spiders known to trick other spiders into thinking that their web contains a struggling insect. When the web-owner hurries over for the kill, she herself becomes the prey. Instead of knowing at birth how to enact a trapped insect, jumping spiders seem to learn how to do so by trial and error. They try out a kaleidoscope of random pluckings and vibrations on the silk of another spider, using their palps and legs, while taking note which signals best lure the owner toward them. The most effective signals will be repeated on future occasions. This tactic allows them to fine-tune their mimicry to any victim species, which is why arachnologists have begun to speak of spider cognition.31

  And why not?

  When in Rome

  To our surprise, chimps turn out to be conformists. Copying others for one’s own benefit is one thing, but wanting to act like everybody else is quite another. It is the foundation of human culture. We discovered this tendency when Vicky Horner presented two separate groups of chimpanzees with an apparatus from which food could be extracted in two different ways. The apes could either poke a stick into a hole to release a grape or use the same stick to lift up a little trap and a grape would roll out. They learned the technique from a model: a pretrained group member. One group saw a lifting model, the other a poking model. Even though we used the same apparatus for both groups, moving it back and forth between them, the first learned to lift, and the second to poke. Vicky had created two distinct cultures, dubbed the “lifters” and the “pokers.”32

  There were exceptions, though. A few individuals discovered both techniques or used a different one than their model had demonstrated. When we retested the chimps two months later, though, most of the exceptions had vanished. It was as if all the apes had settled on a group norm, following the rule “Do what everyone else is doing regardless of what you found out by yourself.” Since we never noticed any peer pressure nor any advantage of one technique over the other, we attributed this uniformity to a conformist bias. Such a bias obviously fits my ideas about imitation guided by a sense of belonging as well as what we know about human behavior. Members of our own species are the ultimate conformists, going so far as to abandoning their personal beliefs if they collide with the majority view. Our openness to suggestion goes well beyond what we found in the chimps, yet it seems related. This is why the conformist label stuck.33

  It is increasingly applied to primate culture, such as by Susan Perry in her fieldwork on capuchin monkeys. Perry’s monkeys have two equally efficient ways of shaking the seeds out of the Luehea fruits that they encounter in the Costa Rican jungle. They can either pound the fruits or rub them on a tree branch. Capuchins are the most vigorous and enthusiastic foragers I know, and most adults develop one technique or the other but not both. Perry found conformism in daughters, who adopted the preferred method of their mothers, but not in sons.34 This sex difference, also known of juvenile chimpanzees learning to fish for termites with twigs, makes sense if social learning is driven by identification with the model. Mothers act as role models for daughters but not necessarily for sons.35

  Conformism is hard to substantiate in the field. There are too many alternative explanations for why one individual might act like another, including genetic and ecological ones. How these issues can be resolved was shown by a large-scale project on humpback whales in the Gulf of Maine in the northeastern United States. In addition to their regular bubble-feeding, in which whales drive fish together with air bubbles, one male invented a new technique. First seen in 1980, this whale would whack the ocean surface with his
fluke to produce a loud noise that clumped the prey even more. Over time this lobtail technique became increasingly common in the population. In the course of a quarter-century, investigators carefully plotted how it spread across six hundred individually recognized whales. They found that whales who had associated with those employing the technique were more likely to use it themselves. Kinship could be ruled out as a factor, because whether a whale had a lobtail-feeding mother hardly mattered. It all boiled down to whom they had encountered while feeding on fish. Since large cetaceans are unsuitable for experiments, this may be as close as we will ever get to proving that a habit spread socially as opposed to genetically.36

  On wild primates, experimental work is rare for different reasons. First of all, these animals are neophobic, and rightly so, because imagine the danger of freely approaching human contraptions, including those set by poachers. Second, fieldworkers generally hate to expose their animals to artificial situations, since their goal is to study them with as little disturbance as possible. Third, they have no control over who participates in an experiment and for how long, thus precluding the kind of tests typically applied to animals in captivity.

  So one has to admire one of the most elegant experiments on conformism on wild monkeys, carried out by the Dutch primatologist Erica van de Waal (no relation).37 Teaming up with Andy Whiten, who has been an engine of cultural studies, van de Waal gave vervet monkeys in a South African game reserve open plastic boxes filled with maize corn. These small grayish monkeys with black faces love corn, but there was a catch: the scientists had manipulated the supply. There were always two boxes with two colors of corn, blue and pink. One color was good to eat whereas the other was laced with aloe, making it disgusting. Depending on which color corn was palatable, and which not, some groups learned to eat blue, and others pink.

  This preference is easily explained by associative learning. But then the investigators removed the distasteful treatment and waited for infants to be born and new males to immigrate from neighboring areas. They watched several groups of monkeys that were supplied with perfectly fine corn of both colors. All adults stubbornly stuck to their acquired preference, however, and never discovered the improved taste of the alternative color. Twenty-six of twenty-seven newborn infants learned to eat only the locally preferred food. Like their mothers, they didn’t touch the other color, even though it was freely available and just as good as the other. Individual exploration was obviously suppressed. The youngsters might even sit on top of the box with the rejected corn while happily feeding on the other type. The single exception was an infant whose mother was so low in rank, and so hungry, that she occasionally tasted the forbidden fruits. Thus, all newborns copied their mothers’ feeding habits. Male immigrants, too, ended up adopting the local color even if they arrived from groups with the opposite preference. That they switched their preference strongly suggests conformism, since these males knew from experience that the other color was perfectly edible. They simply followed the adage “When in Rome . . .”

  These studies prove the immense power of imitation and conformism. It is not a mere extravagance that animals occasionally engage in for trivial reasons—which, I hate to say, is how animal traditions have sometimes been derided—but a widespread practice with great survival value. Infants who follow their mother’s example of what to eat and what to avoid obviously stand a better chance in life than infants who try to figure out everything on their own. The idea of conformism among animals is increasingly supported for social behavior as well. One study tested both children and chimpanzees on generosity. The goal was to see if they were prepared to do a member of their own species a favor at no cost to themselves. They indeed did so, and their willingness increased if they themselves had received generosity from others—any others, not just their testing partner. Is kind behavior contagious? Love begets love, we say, or as the investigators put it more dryly, primates tend to adopt the most commonly perceived responses in the population.38

  The same can be concluded from an experiment in which we mixed two different macaques: rhesus and stumptail monkeys. Juveniles of both species were placed together, day and night, for five months. These macaques have strikingly different temperaments: rhesus are a quarrelsome, nonconciliatory bunch, whereas stumptails are laid-back and pacific. I sometimes jokingly call them the New Yorkers and Californians of the macaque world. After a long period of exposure, the rhesus monkeys developed peacemaking skills on a par with those of their more tolerant counterparts. Even after separation from the stumptails, the rhesus showed nearly four times more friendly reunions following fights than is typical of their species. These new and improved rhesus monkeys confirmed the power of conformism.39

  One of the most intriguing sides of social learning—defined as learning from others—is the secondary role of reward. While individual learning is driven by immediate incentives, such as a rat learning to press a lever to obtain food pellets, social learning doesn’t work this way. Sometimes conformism even reduces rewards—after all, the vervet monkeys missed out on half of the available food. We once conducted an experiment in which capuchin monkeys watched a monkey model open one of three differently colored boxes. Sometimes the boxes contained food, but at other times they were empty. It didn’t matter: the monkeys copied the model’s choices regardless of whether there was any reward.40

  There are even examples of social learning in which the benefits, instead of going to the performer, go to someone else. At the Mahale Mountains in Tanzania, I regularly saw a chimpanzee walk up to another, vigorously scratch the other’s back with his or her fingernails, then settle down to groom the other. In between the grooming, more scratching might follow. This behavior has been known for a long time and has thus far been reported for only one other field site. It is a locally learned tradition, but here’s the rub: when one scratches oneself, it is usually due to itching, and the act brings instant relief. In the case of the social scratch, however, the performer does not feel relief—the recipient does.41

  Primates occasionally learn habits from others that do pay off, such as when chimpanzee youngsters learn to crack nuts with stones. But even then things are not as simple as they appear. Sitting next to their nut-cracking moms, infant chimps are total klutzes. They put nuts on top of stones, stones on top of nuts, and push them all together in a heap only to rearrange them over and over. They gain nothing from this playful activity. They also hit nuts with a hand, or stamp them hard with a foot, which fails to crack anything. Palm and panda nuts are far too tough for them. Only after three years of futile efforts do young chimps have enough coordination and strength to break open their first nut with a pair of stones, but they still have to wait until they are six or seven to reach adult skill levels.42 Since they utterly fail at this task for so many years in a row, it is unlikely that food is the incentive. They may even experience negative consequences, such as smashed fingers. Yet young chimps happily persist, inspired by the example of their elders.

  How little rewards matter is also evident from habits that lack benefits. In our own species, we have fads such as wearing a baseball cap backward or pants that hang low enough to impede locomotion. But in other primates, too, we find seemingly useless fashions and habits. A nice example is the N-family in a group of rhesus monkeys that I observed long ago at the Wisconsin Primate Center. This matriline was headed by an aging matriarch, Nose, all of whose offspring had names starting with the same letter, such as Nuts, Noodle, Napkin, Nina, and so on. Nose had developed the odd routine of drinking from a water basin by dipping her entire underarm into it, then licking her hand and the hair on her arm. Amusingly, all her offspring, and later her grandchildren adopted the exact same technique. No other monkeys in the troop, or any other that I knew, drank like this, yet there was absolutely no advantage to it. It did not allow the N-family to access anything that other monkeys had no access to.

  Or take the way chimpanzees sometimes develop local dialects, such as the excited grunts u
ttered while snacking on tasty food. These grunts differ not only from group to group but also per food type, such as a particular grunt heard only while they eat apples. When the Edinburgh Zoo introduced chimpanzees from a Dutch zoo to its residents, it took those others three years to get socially integrated. Initially, the newcomers uttered different grunts while eating apples, but by the end they converged on the same grunts as the locals. They had adjusted their calls so that they sounded more like those of the residents. While the media hyped this finding by saying that Dutch chimps had learned to speak Scottish, it was more like picking up an accent. The bonding between individuals of different backgrounds had resulted in conformism, even though chimps are not particularly known for vocal flexibility.43

  Clearly, social learning is more about fitting in and acting like others than about rewards. This is why my book on animal culture was entitled The Ape and the Sushi Master. I chose this title partly to honor Imanishi and the Japanese scientists who gave us the animal culture concept, but also because of a story I had heard about how apprentice sushi masters learn their trade. The apprentice slaves in the shadow of the master of an art requiring rice of the right stickiness, precisely cut ingredients, and the eye-catching arrangements for which Japanese cuisine is famous. Anyone who has ever tried to cook rice, mix it with vinegar, and cool it off with a handheld fan so as to mold fresh rice balls in one’s hands knows how complex a skill it is, and it is only a small part of the job. The apprentice learns mostly through passive observation. He washes the dishes, mops the floor, bows to the clients, fetches ingredients, and in the meantime follows from the corners of his eyes, without ever asking a question, everything the sushi master does. For three years he watches without being allowed to make actual sushi for the patrons of the restaurant: an extreme case of exposure without practice. He is waiting for the day when he will be invited to make his first sushi, which he will do with remarkable dexterity.