Are We Smart Enough to Know How Smart Animals Are
Page 8
While corvids always impress us, sheep seem to go a step further in that they remember one another’s faces. British scientists led by Keith Kendrick taught sheep the difference between twenty-five pairs of their own species’ faces by rewarding a choice for one face and not for the other. To us, all these faces look eerily alike, but the sheep learned and retained the twenty-five differences for up to two years. In doing so, they used the same brain regions and neural circuits as humans, with some neurons responding specifically to faces and not to other stimuli. These special neurons were activated if the sheep saw pictures of companions that they remembered—they actually called out to these pictures as if the individuals were present. Publishing their study under the subtitle “sheep are not so stupid after all”—a title to which I object, since I don’t believe in stupid animals—the investigators put the face-recognition ability of sheep on a par with that of primates and speculated that a flock, which to us looks like an anonymous mass, is in fact quite differentiated. This also means that mixing flocks, as is sometimes done, may cause more distress than we realize.
Having made primate chauvinists look sheepish, science piled it on with wasps. The northern paperwasp, common in the American Midwest, has a highly structured society with a hierarchy among its founding queens, who are dominant over all workers. Given the intense competition, each wasp needs to know her place. The alpha queen lays most eggs, followed by the beta queen, and so on. Members of the small colony are aggressive to outsiders as well as to females whose facial markings have been altered by experimenters. They recognize one another by strikingly different patterns of yellow and black on every female’s face. The American scientists Michael Sheehan and Elizabeth Tibbetts tested individual recognition and found it to be as specialized as that of primates and sheep. The wasps distinguish their own species’ mugs far better than other visual stimuli, and they also outperform a closely related wasp that lives in colonies founded by a single queen. These wasps hardly have a hierarchy and have far more homogeneous faces. They don’t need individual recognition.10
If face recognition has evolved in such disparate pockets of the animal kingdom, one wonders how these capacities connect. Wasps do not have the big brains of primates and sheep—they have minuscule sets of neural ganglia—hence they must be doing it in a different manner. Biologists never tire of stressing the distinction between mechanism and function: it is very common for animals to achieve the same end (function) by different means (mechanism). Yet with respect to cognition, this distinction is sometimes forgotten when the mental achievements of large-brained animals are questioned by pointing at “lower” animals doing something similar. Skeptics delight in asking “If wasps can do it, what’s the big deal?” This race to the bottom has given us trained pigeons hopping onto little boxes to disparage Köhler’s experiments on apes and the holding up of intelligence outside the primate order to cast doubt on mental continuity between humans and other hominoids.11 The underlying thought is that of a linear cognitive scale, and the argument that since we rarely assume complex cognition in “lower” animals, there is no reason to do so in “higher” ones.12 As if there were only one way to achieve a given outcome!
Evolutionary science distinguishes between homology (the traits of two species derive from a common ancestor) and analogy (similar traits evolved independently in two species). The human hand is homologous with the bat’s wing since both derive from the vertebrate forelimb, as is recognizable by the shared arm bones and five phalanges. The wings of insects, on the other hand, are analogous to those of bats. As products of convergent evolution, they serve the same function but have a different origin.
This is not the case. Nature abounds with illustrations to the contrary. One that I know firsthand is a pair-bonding Amazonian cichlid, the discus fish, that has achieved the equivalent of mammalian nursing. Once the fry have absorbed the egg yolk, they gather along the flanks of Mom and Dad to nibble mucus off their bodies. The breeding pair secretes extra mucus for this purpose. The young enjoy both nutrition and protection for about a month until they are “weaned” by parents who now turn away each time they approach.13 No one would use these fish to make a point about the complexity or simplicity of mammalian nursing for the obvious reason that the mechanisms are radically different. All that they share is the function of feeding and raising the young. Mechanism and function are the eternal yin and yang of biology: they interact and intertwine, yet there is no greater sin than confusing the two.
To understand how evolution works its magic across the evolutionary tree, we often invoke the twin concepts of homology and analogy. Homology refers to shared traits derived from a common ancestor. Thus, the human hand is homologous with the wing of a bat, since both derive from an ancestral forelimb and carry the exact same number of bones to prove it. Analogies, on the other hand, arise when distant animals independently evolve in the same direction, known as convergent evolution. The parental care of the discus fish is analogous to mammalian nursing but certainly not homologous, since fish and mammals do not share an ancestor that did the same. Another example is how dolphins, ichthyosaurs (extinct marine reptiles), and fish all have strikingly similar shapes owing to an environment in which a streamlined body with fins serves speed and maneuverability. Since dolphins, ichthyosaurs, and fish did not share an aquatic ancestor, their shapes are analogous. We can apply the same line of thought to behavior. The sensitivity to faces in wasps and primates came about independently, as a striking analogy, based on the need to recognize individual group mates.
Convergent evolution is incredibly powerful. It has equipped both bats and whales with echolocation, both insects and birds with wings, and both primates and opossums with opposable thumbs. It has also produced spectacularly similar species in distant geographic regions, such as the armored bodies of armadillos and pangolins, the prickly defense of hedgehogs and porcupines, and the predatory weaponry of the Tasmanian tiger and the coyote. There is even a primate, the aye-aye of Madagascar, that looks like E.T. with an extremely elongated middle finger (to tap for hollow spots and extract grubs from wood), a trait that it shares with a small marsupial, the long-fingered triok of New Guinea. These species are genetically miles apart, yet they have evolved the same functional solution. We should not be surprised therefore to find similar cognitive and behavioral traits in species that are eons and continents apart. Cognitive rippling is common precisely because it isn’t bound by the evolutionary tree: the same capacity may pop up almost anywhere it is needed. Instead of taking this as an argument against cognitive evolution, as some have done, it perfectly fits the way evolution works through either common descent or adaptation to similar circumstances.
A prime example of convergent evolution is the use of tools.
Redefining Man
As soon as an ape sees something attractive yet out of reach, he starts to cast about for a bodily extension. An apple floats in the moat around the zoo island: the ape takes one glance at the fruit before racing around in search of a suitable stick or a few stones that he can throw behind it so that it will float toward him. He distances himself from his goal in order to reach it—an illogical thing to do—while carrying a search image of what tool might work best. He is in a hurry, because if he doesn’t return fast enough, someone else will beat him to the prize. If, on the other hand, his goal is to eat fresh green leaves from a tree, the required tool is quite different: something sturdy to climb on. He may work for half an hour to drag and roll a heavy loose tree stump in the direction of the one tree on the island that has a low side branch. The whole reason he needs a tool is to get across the electric wire around the tree. Before making the actual attempt, he has figured out that the low branch will come in handy. I have even seen apes check the hot wires with the hair on the back of their wrist, hand bent inward, barely touching it, but enough to know if the power is on. If it is off, obviously no tool will be needed, and the foliage is fair game.
Apes do not just search for tool
s for specific occasions; they actually fabricate them. When the British anthropologist Kenneth Oakley, in 1957, wrote Man the Toolmaker, which claimed that only humans make tools, he was well aware of Köhler’s observations of Sultan fitting sticks together. But Oakley refused to count this as tool manufacture, since it was done in reaction to a given situation rather than in anticipation of an imagined future. Even today some scholars dismiss ape tools by stressing how human technology is embedded in social roles, symbols, production, and education. A chimpanzee cracking nuts with rocks doesn’t qualify; nor, I suspect, does a farmer picking his teeth with a twig. One philosopher even felt that since chimpanzees don’t need their so-called tools, it remains a feeble comparison.14
One of the most complex tool skills is the cracking of tough nuts with rocks. A wild female chimpanzee selects an anvil stone and finds a hammer that fits her hand to open a nut, while her son watches and learns. Only by the age of six will he reach adult proficiency.
I feel like recalling my know-thy-animal rule here, according to which we can safely dismiss a philosopher who thinks that wild chimpanzees sit there pounding and pounding hard nuts with rocks, an average of thirty-three blows per consumed kernel, for generation after generation, for no good reason at all. During peak season, chimpanzees at some field sites spend close to 20 percent of their waking hours fishing with twigs for termites or cracking nuts between rocks. It is estimated that they gain nine times as many kilocalories of energy from this activity as they put into it.15 Moreover, the Japanese primatologist Gen Yamakoshi found that nuts serve as fallback foods when the apes’ main nutrition—seasonal fruits—is scarce.16 Another fallback is palm pith, which is obtained through “pestle pounding.” High up in a tree, a chimpanzee stands bipedally at the edge of the tree crown, pounding the top with a leaf stalk, thus creating a deep hole from which fiber and sap can be collected. In other words, the survival of chimpanzees is quite dependent on tools.
Ben Beck gave us the best-known definition of tool use, of which the short version goes as follows: “the external deployment of an unattached environmental object to alter more efficiently the form, position, or condition of another object.”17 Though imperfect, this definition has served the field of animal behavior for decades.18 Tool manufacture can then be defined as the active modification of an unattached object to make it more effective in relation to one’s goal. Note that intentionality matters a great deal. Tools are brought in from a distance and modified with a goal in mind, which is the reason traditional learning scenarios, which revolve around accidentally discovered benefits, have such trouble explaining this behavior. If you see a chimpanzee strip the side branches off a twig to make it right for ant fishing, or collect a fistful of fresh leaves and chew them into a spongelike clump to absorb water from a tree hole, it is hard to miss the purposefulness. By making suitable tools out of raw materials, chimpanzees are exhibiting the very behavior that once defined Homo faber, man the creator. This is why the British paleontologist Louis Leakey, when he first heard about such behavior from Goodall, wrote her back, “I feel that scientists holding to this definition are faced with three choices: They must accept chimpanzees as man, they must redefine man, or they must redefine tools.”19
After the many observations of chimpanzee tool use in captivity, seeing tool use in the wild by the same species did perhaps not come as a surprise, yet its discovery was crucial since it could not be explained away by human influence. Moreover, wild chimps not only use and make tools, but they learn from one another, which allows them to refine their tools over generations. The result is more sophisticated than anything we know in zoo chimps. A good example are the toolkits, which can be so complex that it is hard to imagine that they were invented in a single step. A typical one was found by the American primatologist Crickette Sanz in the Goualougo Triangle, Republic of Congo, where a chimpanzee may arrive with two different sticks at a particular open spot in the forest. It is always the same combination: one is a stout woody sapling of about a meter long, while the other is a flexible slender herb stem. The chimp then proceeds to deliberately drive the first stick into the ground, working it with both hands and feet the way we do with a shovel. Having made a sizable hole to perforate an army ant nest deep under the surface, she pulls out the stick and smells it, then carefully inserts her second tool. The flexible stem captures bite-happy insects that she pulls up and eats, dipping regularly into the nest below. Apes often climb off the ground, moving onto tree buttresses, to avoid the nasty bites of colony defenders. Sanz collected more than one thousand such tools, which shows how common the perforator-dipping combination is.20
More elaborate toolkits are known for chimpanzees in Gabon hunting for honey. In yet another dangerous activity, these chimps raid bee nests using a five-piece toolkit, which includes a pounder (a heavy stick to break open the hive’s entrance), a perforator (a stick to perforate the ground to get to the honey chamber), an enlarger (to enlarge an opening through sideways action), a collector (a stick with a frayed end to dip into honey and slurp it off), and swabs (strips of bark to scoop up honey).21 This tool use is complicated since the tools are prepared and carried to the hive before most of the work begins, and they will need to be kept nearby until the chimp is forced to quit due to aggressive bees. Their use takes foresight and planning of sequential steps, exactly the sort of organization of activities often emphasized for our human ancestors. At one level chimpanzee tool use may seem primitive, as it is based on sticks and stones, but on another level it is extremely advanced.22 Sticks and stones are all they have in the forest, and we should keep in mind that also for the Bushmen the most ubiquitous instrument is the digging stick (a sharpened stick to break open anthills and dig up roots). The tool use of wild chimpanzees by far exceeds what was ever held possible.
Chimpanzees use between fifteen and twenty-five different tools per community, and the precise tools vary with cultural and ecological circumstances. One savanna community, for example, uses pointed sticks to hunt. This came as a shock, since hunting weapons were thought to be another uniquely human advance. The chimpanzees jab their “spears” into a tree cavity to kill a sleeping bush baby, a small primate that serves as protein source for females unable to run down monkeys the way males do.23 It is also well known that chimpanzee communities in West Africa crack nuts with stones, a behavior unheard of in East African communities. Human novices have trouble cracking the same tough nuts, partly because they do not have the same muscle strength as an adult chimpanzee, but also because they lack the required coordination. It takes years of practice to place one of the hardest nuts in the world on a level surface, find a good-sized hammer stone, and hit the nut with the right speed while keeping one’s fingers out of the way.
The Japanese primatologist Tetsuro Matsuzawa tracked the development of this skill at the “factory,” an open space where apes bring their nuts to anvil stones and fill the jungle with a steady rhythm of banging noise. Youngsters hang around the hardworking adults, occasionally pilfering kernels from their mothers. This way they learn the taste of nuts as well as the connection with stones. They make hundreds of futile attempts, hitting the nuts with their hands and feet, or aimlessly pushing nuts and stones around. That they still learn the skill is a great testament to the irrelevance of reinforcement, because none of these activities is ever rewarded until, by about three years of age, the juvenile starts to coordinate to the point that a nut is occasionally cracked. It is only by the age of six or seven that their skill reaches adult level.24
When it comes to tool use, chimps always catch the limelight, but there are three other great apes—bonobos, gorillas, and orangutans—that, together with chimps, us, and the gibbons, make up the Hominoid family. Not to be confused with monkeys, Hominoids are large, flat-chested primates without tails. Within this family, we are closest to chimps and bonobos, both of which are genetically nearly identical to us. Naturally, there is heated debate about what the minuscule-sounding 1.2 percent DNA diff
erence between us and them exactly means, but that we are close family is not in doubt. In captivity, the orangutan is an absolute master tool user, dexterous enough to tie knots into loose shoelaces, and to construct instruments. One young male was seen to join three sticks, which he had first sharpened, into two tubes to build a five-section pole to knock down suspended food.25 Being notorious escape artists, orangs may dismantle their cage so patiently, from day to day and week to week, while keeping dislodged screws and bolts out of sight, that keepers fail to notice what they are doing until it’s too late. In contrast, until recently all we knew about wild orangs was that they sometimes scratched their butt with a stick or held a leafy branch over their head during rain. How could a species that is so talented offer so little evidence of tool use in the wild? The inconsistency was resolved when, in 1999, the tool technology of orangutans in a Sumatran peat swamp came to light. These orangs extract honey from bee nests with twigs and use short sticks to remove the seeds embedded in the stinging hairs of neesia fruits.26
The other ape species, too, are perfectly capable of tool use, and we have already laid to rest the view that gibbons lack this capacity.27 But reports from the wild remain meager to nonexistent, sometimes suggesting that only chimps are proficient tool users. We see glimpses, such as when gorillas preventively disarm poacher snares, which requires a grasp of basic mechanics, or traverse deep water. When elephants had dug a new water hole in a swampy forest in the Republic of Congo, the German primatologist Thomas Breuer saw a female gorilla, Leah, try to wade across. She stopped when she was waist-deep into it, however—apes hate swimming. Leah returned to shore to pick up a long branch to gauge the water’s depth. Feeling around with her stick, she walked bipedally far into the pool before retracing her steps to return to her wailing infant. This example highlights the shortcomings of Beck’s classical definition, because even though Leah’s stick altered neither anything in the environment nor her own position, it did serve as a tool.28