Are We Smart Enough to Know How Smart Animals Are

by Frans de Waal

  Chimpanzees are recognized as the most versatile primate tool user apart from us, but this heralded position has been challenged. The challenge did not come from any hominoid but from a small South American monkey. Brown capuchin monkeys have been known for centuries as organ-grinders and more recently as trained helping hands for quadriplegics. They are extremely manipulative and particularly good at tasks that tap into their tendency to smash and bang things. Having had a colony of these monkeys for decades, I know that almost anything you hand them (a piece of carrot, an onion) is going to be pounded to mush on the floor or against the wall. In the wild, they pound oysters for a long time until the mollusk relaxes its muscle so they can pry it open. During the fall, our monkeys in Atlanta collected so many fallen hickory nuts from nearby trees that we’d hear frantic banging sounds the whole day in our office adjacent to the monkey area. It was a happy sound, because capuchins seem to be in their best mood when they are doing things. Not only did they try to break open the nuts, they also employed hard objects (a plastic toy, a block of wood) to smash them with. About half the members of one group learned to do so, whereas the second group never invented the technique despite having the same nuts and tools. This group obviously consumed fewer nuts.

  Capuchins’ natural predisposition to be persistent pounders sets them up for nut cracking in the field. A Spanish naturalist first reported it five centuries ago, and more recently an international team of scientists found dozens of cracking sites in the Tietê Ecological Park and other sites in Brazil.29 At one site, capuchins eat the pulp of a large fruit, after which they drop its seeds to the ground. They return a couple days later to collect those seeds, which by then have dried out and are often infested with larvae, which the monkeys are fond of. Traveling with the seeds stuffed in their hands, mouth, and (prehensile) tail in search of a hard surface, such as a large rock, the monkeys would get a smaller stone to pound the seeds with. While these stones are about the same size as those used by chimps, the capuchins are only about the size of a small cat, so their hammers weigh about one-third of their body! Literally acting as heavy equipment operators, they lift them high above their heads to get a good hit. When the tough seeds are cracked, the larvae are there for the picking.30

  Capuchin nut cracking thoroughly upset the evolutionary narrative that had been woven around humans and apes. According to this story, we are not the only ones who knew a Stone Age: our closest relatives still live in one. To stress this point, a “percussive stone technology” site (including stone assemblies and the remains of smashed nuts) was excavated in a tropical forest in Ivory Coast, where chimpanzees must have been opening nuts for at least four thousand years.31 These discoveries led to a human-ape lithic culture story that fit together nicely, tying us to our close relatives.

  This is why the discovery of similar behavior in a more distant relative, such as the capuchin monkey—equipped with tails by which they can hang!—was met with surprise and initial grumbling. The monkeys didn’t fit. The more we learned, however, the more the nut cracking by capuchins in Brazil began to resemble that of chimpanzees in West Africa. Yet capuchins belong to the neotropical monkeys, a distant group that split off 30 to 40 million years ago from the rest of the primate order. Perhaps the similar tool use was a case of convergent evolution, since both chimps and capuchins are extractive foragers. They break things open, destroy outer shells, and smash things to pulp in order to eat, which might be the context in which their high intelligence evolved. On the other hand, since both are large-brained primates with binocular vision and manipulative hands, there is an undeniably evolutionary connection. The distinction between homology and analogy is not always as clear as we’d like it to be.

  To complicate matters, the tool use of capuchins and chimpanzees may not be cognitively at the same level. Over many years of working with both species, I have formed a distinct impression of how they go about their business, which I’ll offer here in everyday language. Chimpanzees, like all the apes, think before they act. The most deliberate ape is perhaps the orangutan, but chimps and bonobos, despite their emotional excitability, also judge a situation before tackling it, weighing the effect of their actions. They often find solutions in their heads rather than having to try things out. Sometimes we see a combination of both, as when they start acting on a plan before it is completely formed, which is of course not unusual in our species either. In contrast, the capuchin monkey is a frenzied trial-and-error machine. These monkeys are hyperactive, hypermanipulative, and afraid of nothing. They try out a great variety of manipulations and possibilities, and once they discover something that works, they instantly learn from it. They don’t mind making tons of mistakes and rarely give up. There is not much pondering and thinking behind their behavior: they are overwhelmingly action-driven. Even if these monkeys often end up with the same solutions as the apes, they seem to get there in an entirely different way.

  While all this may be a gross simplification, it is not without experimental support. An Italian primatologist, Elisabetta Visalberghi, has spent a lifetime studying the tool use of brown capuchins at her facility adjoining the Rome Zoo. In one illuminating experiment, a monkey faced a horizontal transparent tube with a peanut visible in the center. The plastic tube was mounted so that the peanut was at monkey eye level. The monkey couldn’t get to it, though, since the tube was too narrow and long. Many objects were available to push the food out, ranging from the most suitable (a long stick) to the least (short sticks, soft flexible rubber). The capuchins made an astonishing number of errors, such as hitting the tube with the stick, vigorously shaking the tube, pushing the wrong material into one end, or pushing short sticks into both ends so that the peanut couldn’t budge. The monkeys learned over time, however, and began to prefer the long stick.

  A brown capuchin monkey (top) inserts a long stick into a transparent tube to push out a peanut. Placed in a regular tube, the peanut may be pushed in either direction to solve the problem. The trap-tube (bottom), by contrast, requires the peanut to be pushed in only one direction, otherwise it will drop into the trap and be lost to the monkey. Monkeys can learn to avoid the trap after many errors, but apes show cause-effect understanding and recognize the solution right away.

  At this point, Visalberghi added an ingenious twist by making a hole in the tube. Now it suddenly mattered which way the peanut was pushed. Pushed toward the hole, the peanut would fall into a small plastic container and be lost to the monkey. Would capuchin monkeys understand the need to stay clear of the trap, and would they do so right away or only after many failed attempts?

  Handing four monkeys a long stick to work on the trap tube, three performed at random, being successful half the time, which they seemed perfectly happy with. But not Roberta, a slender young female, who kept trying and trying. She’d push the stick into the left end of the tube, then race around to see how it and the peanut looked from the right end. Then she’d switch sides, inserting the stick into the right end, only to race around to peek into the tube from the left. She kept going back and forth, sometimes failing, sometimes succeeding, but in the end becoming quite successful.

  How had Roberta solved the problem? The investigators concluded that she followed a simple rule of thumb: insert the stick into the end of the tube farthest away from the reward. This way the peanut could be pushed out without having to cross the trap. They tested it out in several ways, one of which was to offer Roberta a new plastic tube without any trap at all. Now she could push the stick whichever way she wanted and be successful. She kept racing around the tube, however, looking for the longest distance from the peanut, insisting on the rule that had been the key to her success. Since Roberta acted as if the trap were still there, she clearly had not paid much attention to how it worked. Visalberghi concluded that monkeys are able to solve the trap-tube task without actually understanding it.32

  This task may look simple, yet is harder than it seems: human children solve it reliably only when they are over thre
e years old. And testing five chimpanzees on the same problem, two of them grasped the cause-effect relation and learned to specifically avoid the trap.33 While Roberta had merely learned which actions led to success, the apes recognized how the trap worked. They were representing the connections between actions, tools, and outcomes in their heads. This is known as a representational mental strategy, which allows solutions before action. This difference may seem minor, since both monkeys and apes solved the problem, but it is actually huge. The level at which apes understand the purpose of tools affords them incredible flexibility. The richness of their technology, the toolkits, and the frequent toolmaking all prove that higher cognition helps. The American primate expert William Mason concluded in the 1970s that evolution has endowed the hominoids’ with a cognition that sets them apart from the other primates, so that an ape is best described as a thinking being.

  The ape structures the world in which it lives, giving order and meaning to its environment, which is clearly reflected in its actions. It is not very illuminating, perhaps, to describe a chimpanzee as “figuring out” how to proceed, while it sits and stares at the problem before it. Certainly such an assertion lacks originality, as well as precision. But we cannot escape the inference that some such process is at work, and that it has a significant effect on the ape’s performance. It seems better to be vaguely correct than positively wrong.34

  Here Come the Crows!

  I first encountered the tube task during a visit to Jigokudani Monkey Park, in Japan, in one of the world’s coldest habitats with native primates. Tourist guides use the task to demonstrate monkey intelligence. At the feeding site next to the river, which attracts snow monkeys from the surrounding montane forest, a horizontal transparent tube was baited with a piece of sweet potato. Rather than wielding a stick like the capuchins, one female snow monkey pushed her small infant into the tube while firmly holding on to its tail. The baby crawled toward the food and grabbed it, only to be quickly withdrawn by its loving mom, who pried the prize from its resistant clutch. Another female collected rocks to throw into one end of the tube, so that the food came out the other end.

  These are macaques, monkeys much closer to us than capuchins. The most spectacular evidence for macaque tool use has been collected by Michael Gumert, an American primatologist. On Piak Nam Yai Island off the coast of Thailand, Gumert found an entire population of long-tailed macaques using stone tools. I am very familiar with this species, having done my dissertation on them. Also known as crab-eating macaques, these smart monkeys are rumored to hang their long tails in water to pull up crabs. I have seen them use their tail almost like a stick to obtain food. Unable to control it as South American primates do—a macaque’s tail is nonprehensile—they grab their tail with one hand and swap food from outside to inside their cage with it.

  Manipulating one’s own body appendage is yet another example that stretches the definition of tool use, but there is no doubt that what Gumert discovered is a well-developed technology. His monkeys on the coast collect stones everyday for two purposes. Bigger stones serve as hammers to pound oysters with blunt force until they break open, revealing a delicious rich food source. Smaller stones are used rather like axes, applying a precision grip and more rapid movements, in order to dislodge shellfish from rocks. During the few hours of ebb tide, both food and tools are abundant, an ideal situation for the invention of this seafood technology. It is testimony to the generalized intelligence of primates, because obviously they evolved in the trees, eating fruits and leaves, but here they were surviving on the beach. After humans, chimps, and capuchins, a fourth primate has entered the Stone Age.35

  Beyond the primates, however, there are quite a few tool-using mammals and birds. Coastal Californians can watch their own floating technology every day among the kelp. The popular furry sea otter swims on his back while using both front paws to smash shellfish against a stone anvil on his chest. He also hammers abalones with a large rock to dislodge them, taking multiple dives to finish this underwater job. A close relative of the otter may possess even more spectacular talents. The honey badger is the star of a viral YouTube video full of expletives to indicate how “badass” this Chuck Norris of the animal kingdom is. The species is even featured on T-shirts emblazoned with “Honey badger don’t care.” This so-called badger is a small carnivore, which actually—like the otter—belongs to the weasel family. While I know of no official reports about their skills, a recent PBS documentary features a rescued honey badger named Stoffel who invents multiple ways to escape from his enclosure at a South African rehabilitation center.36 Assuming that what we see is not a trained trick, he outwits his human caretakers at every turn and displays the sort of insight for his Houdini act that one might expect from an ape, not a honey badger. The documentary shows Stoffel leaning a rake against the wall and claims that he once piled up large stones against it to escape. After all the stones were removed from his enclosure, he apparently constructed a heap of mud balls for the same purpose.

  Even though all this is most impressive and begs for further investigation, the greatest challenge to the supremacy of primates has come, not from other mammals, but from a flock of squawking and cawing birds that landed right in the midst of the tool debate. They caused about as much mayhem as they did in that Hitchcock movie.

  During the quiet hours in his pet store, my grandfather patiently trained goldfinches to pull a string. This particular finch is known in Dutch as a puttertje, a name that refers to the drawing of water from a well. Males that could both sing and draw would fetch a high price. For centuries, these little colorful birds were kept in homes with a chain around their leg, pulling a thimble up from a glass so as to fetch their own drinking water. One such finch is featured in the seventeenth-century Dutch painting central to Donna Tartt’s novel The Goldfinch. Of course, we don’t keep these birds anymore, at least not in this cruel fashion, but their traditional trick is very similar to the one that, in 2002, gave us Betty the crow.

  In an aviary at Oxford University, Betty was trying to pull a little bucket out of a transparent vertical pipe. In the bucket was a small piece of meat, and next to the pipe were two tools for her to choose from. One was a straight wire, the other a hooked one. Only with the latter could Betty get a hold of the bucket’s handle. After her companion stole the hooked wire, however, she faced the task with an inappropriate tool. Undeterred, Betty used her beak to bend the straight wire into a hook so as to pull the bucket from the tube. This remarkable feat was a mere anecdote until perceptive scientists systematically investigated it with new tools. In subsequent tests, Betty received only straight wires, which she kept subjecting to her remarkable bending act.37 Apart from dispelling the “birdbrain” notion with which birds are unfairly saddled, Betty achieved instant fame by giving us the first laboratory proof of toolmaking outside the primate order. I add “laboratory,” because Betty’s species in the wild, in the Southwest Pacific, was already known for tool crafting. New Caledonian crows spontaneously modify branches until they have a little wooden hook to fish grubs out of crevices.38

  Inspired by an Aesop fable, crows have been tested to see if they will throw stones into a tube filled with water to bring floating rewards within reach. They do.

  The Ancient Greek poet Aesop may have had an inkling of these talents given his fable The Crow and the Pitcher. “A Crow, half-dead with thirst,” so the fabulist went, “came upon a Pitcher.” There wasn’t enough water in the pitcher for the crow to be able to drink it. He tried to reach in with his beak, but the water level was too low. “Then a thought came to him,” as Aesop put it, “and he took a pebble and dropped it into the Pitcher.” Many more pebbles followed until the water had risen enough for a drink. It seems an unlikely feat, but it has now been replicated in the lab. The first was an experiment on rooks, a corvid that in the wild does not use any tools. The rooks were presented with a vertical water-filled tube with a floating mealworm just out of reach. The water level would have to be
raised if the rook were to reach the delicacy. The same experiment was carried out with New Caledonian crows, known as real tool experts. True to the dictum that necessity is the mother of invention, and confirming Aesop’s story several millennia later, both crow species successfully solved the floating worm puzzle by using pebbles to raise the water level in the tube.39

  Let me add some caution, though, because it is unclear how insightful this solution was. For one thing, all the birds had been pretrained using a slightly different task. They had received ample rewards for plunging stones into a tube. Moreover, while they were facing the tube with the mealworm, stones had been conveniently placed right next to it. The experimental setup strongly suggested the solution, therefore. Imagine that Köhler had taught his chimps to stack boxes! We would never have heard of him, as it would have undermined any claims of insight. In the course of testing, the crows did learn that large stones work better than small ones, and that there is no point dropping stones into a pipe filled with sawdust. Rather than working these answers out in their minds, however, it may have been a matter of fast learning. Perhaps they noticed that adding stones to water brought the mealworm closer, which led them to persist.40

  When we recently presented our chimpanzees with a floating peanut task, a female named Liza solved it right away, adding water to a plastic tube. After some vigorous but ineffective kicking and shaking of the tube, Liza abruptly turned around, went to the drinker to fill her mouth, and returned to the tube to add water. She made several more trips to the drinker before she got the peanut at the right level to reach it with her fingers. Other chimps were less successful, but one female tried to pee into the tube! She had the right idea even though the execution was flawed. I have known Liza all her life and am sure that this problem was brand new to her.